Subset (5)

F1000Journal · 8.11k rows

Split (1)

data · 8.11k rows

link stringlengths 41 45	date stringlengths 9 9	paper dict	reviews listlengths 1 6	version int64 1 5	main stringlengths 38 42
https://f1000research.com/articles/7-1505/v1	20 Sep 18	{ "type": "Software Tool Article", "title": "Feature selection with the R package MXM", "authors": [ "Michail Tsagris", "Ioannis Tsamardinos", "Ioannis Tsamardinos" ], "abstract": "Feature (or variable) selection is the process of identifying the minimal set of features with the highest predictive performance on the target variable of interest. Numerous feature selection algorithms have been developed over the years, but only few have been implemented in R as a package. The R package MXM is such an example, which not only offers a variety of feature selection algorithms, but has unique features that make it advantageous over its competitors: a) it contains feature selection algorithms that can treat numerous types of target variables, including continuous, percentages, time to event (survival), binary, nominal, ordinal, clustered, counts, left censored, etc; b) it contains a variety of regression models to plug into the feature selection algorithms; c) it includes an algorithm for detecting multiple solutions (many sets of equivalent features); and d) it includes memory efficient algorithms for high volume data, data that cannot be loaded into R. In this paper we qualitatively compare MXM with other relevant packages and discuss its advantages and disadvantages. We also provide a demonstration of its algorithms using real high-dimensional data from various applications.", "keywords": [ "Feature selection", "algorithms", "R package", "computational efficiency" ], "content": "Introduction\n\nGiven a target (response or dependent) variable Y of n measurements and a set X of p features (predictor or independent variables) the problem of feature (or variable) selection (FS) is to identify the minimal set of features with the highest predictability on the target variable (outcome) of interest. Why should researchers and practitioners perform FS? For a variety of reasons1: a) many features may be expensive (and/or unnecessary) to measure, especially in the clinical and medical domains; b) FS may result in more accurate models (of higher predictability) by removing noise while treating the curse-of-dimensionality; c) the produced parsimonious models are computationally cheaper and easier to understand and interpret; d) future experiments can benefit from prior feature selection tasks and provide more insight into the problem of interest, its characteristics and structure.\n\nR contains thousands of packages, but only a small portion of them are dedicated to the task of FS, yet offering limited or narrow capabilities. For example, packages that accept few or specific types of target variables (e.g. binary and multi-class only). This leaves many types of target variables, for example percentages, left censored, positive valued, matched case-control data, etc., untreated. The availability of regression models for some types of data is rather small. Count data is such an example, for which Poisson regression is the only model considered in nearly all R packages. Most algorithms including statistical tests offer limited statistical tests, e.g. likelihood ratio test only. Almost all available FS algorithms are devised for large sample sized data, thus they cannot be used in many biological settings where the number of observations rarely (or never in some cases) exceeds 100, but the number of features is in the order of tens of thousands. Finally, some packages are designed for high volume data1 only.\n\nIn this paper we present MXM2; an R package that overcomes the above shortcomings. It contains many FS algorithms2, which can handle many and diverse types of target variables, while offering a pool of regression models to choose from and employ. There is a plethora of statistical tests (likelihood-ratio, Wald, permutation based) and information criteria (BIC and eBIC) to plug into the FS algorithms. Algorithms that work with small (and large) sample sized data, algorithms that have been customized for high volume data, and an algorithm that returns multiple sets of statistically equivalent features are some of the key characteristics of MXM.\n\nOver the next sections, a brief qualitative comparison of MXM with other packages available on CRAN and Bioconductor is presented, its (dis)advantages are discussed, its FS algorithms and related functions are mentioned. Finally a demonstration takes place, applying some FS algorithms available in MXM on real high dimensional data.\n\n\nThe R package MXM\n\nWhen searching for FS packages on CRAN and Bioconductor repositories using the keywords \"feature selection\", \"variable selection\", \"selection\", \"screening\" and \"LASSO\", we detected 184 R packages until the 7th of May 20183. Table 1 shows the frequency of the target variable types those packages accept, while Figure 1 shows the frequency of R packages whose FS algorithms can treat at least one type of target variable, of those presented in Table 1. Table 2 presents the frequency of pairwise types of target variables offered in R packages and Table 3 contains information on packages allowing for less frequent regression models. Most packages offer FS algorithms that are oriented towards specific types of target variables, methodology and regression models, offering at most 3-4 options. Out of these 184 packages, 65 (35.32%) offer LASSO type FS algorithms, while 19 (10.32%) address the problem of FS from the Bayesian perspective. Only 2 (1.08%) R packages treat the case of FS with multiple datasets while only 4 (2.17%) packages are devised for high volume data.\n\nThe percentage-wise number appears inside the parentheses.\n\nThe horizontal axis shows the number of types (any combinations) of target variables from Table 1. For example, there 95 R packages that can handle only 1 type (any type) of target variable, 41 packages that can handle any 2 types of target variables, while MXM is the only one that handles all of them.\n\nThere are 108 packages which handle binary target variables, 59 packages offering algorithms for binary and continuous target variables and only one package handling ordinal and nominal target variables, etc.\n\nThe percentage-wise number appears inside the parentheses.\n\nTable 4 summarizes the types of target variables treated by MXM’ FS algorithms along with the appropriate regression models that can be employed. The list is not exhaustive, as in some cases the type of the predictor variables (continuous or categorical) affects the decision of using a regression model or a test (Pearson and Spearman for continuous and G2 test of independence for categorical). With percentages for example, MXM offers numerous regression models to plug into its FS algorithms: beta regression, quasi binomial regression or any linear regression model (robust or not) after transforming the percentages using the logistic transformation. For repeated measurements (correlated data), there are two options offered, the GLMM and GEE methodologies which can also be used with various types of target variables, not mentioned here. We emphasize that MXM is the only package that covers all types of response variables mentioned on Table 1, many types of which are not available in any other FS package, such as left censored data for example. MXM also covers 3 out 4 cases that appear on Table 3.\n\nMost of the currently available FS algorithms in the MXM package have been developed by the creators and authors of the package. These algorithms have been tested and compared with other state-of-the-art algorithms under different scenarios and types of data.\n\nIAMB3 was on par with or outperforming competing machine learning algorithms, when both the target variable and features are categorical. MMPC and MMMB algorithms4 were tested in the context of BN learning showing great success with MMPC shown to achieve excellent false positive rates5. MMPC was also used as the basis of MMHC6, a prototypical algorithm for learning the structure of a Bayesian network which outperformed all other Bayesian network learning algorithms with categorical data. For time-to-event and nominal categorical target variable, MMPC 7, and 8 respectively, outperformed or was on par with LASSO and other FS algorithms. SES was contrasted against LASSO2 with continuous, binary and survival target variables, resulting in similar conclusions as before. With temporal and time-course data, SES9 outperformed the LASSO algorithm10 both in predictive performance and computational efficiency. FBED11 was compared with LASSO for the task of binary classification with sparse data exhibiting performance similar to that of LASSO. gOMP, a generalization of OMP12–14, has not been publicly tested, but our anecdotal experiments have showed very promising results, achieving similar or better performance, while enjoying higher computational efficiency than LASSO.\n\nThe main advantage of MXM is that all FS algorithms accept numerous and diverse types of target variables. MMPC, SES and FBED treat all types of target variables presented in Table 4, while gOMP handles fewer types4.\n\nMXM is the only R package that offers many different regression models to be employed by the FS algorithms, even for the same type of response variable, such as Poisson, quasi Poisson, negative binomial and zero inflated Poisson regression for count data. For repeated measurements, the user has the option of using GLMM or the GEE methodology (the latter with more options in the correlation structure) and for time-to-event data, Cox, Weibull and exponential regression models are the available options.\n\nA range of statistical tests and methodologies to select the features is offered. Instead of the usual log-likelihood ratio test, the user has the option to use the Wald test or produce a p-value based on permutations. The latter is useful and advised when the sample size is small, emphasizing the need for use of MMPC and SES, both of which are designed for small sample sized datasets. FBED on the other hand gives the option of using information criteria, BIC15 or eBIC16, instead of the log-likelihood ratio test.\n\nStatistically Equivalent Signatures (SES)2,17 builds upon the ideas of MMPC and returns multiple (statistically equivalent) sets of predictor variables, making it one one of the few FS algorithms suggested in the literature, and available in hrefhttps://cran.r-project.org/CRAN, with this trait. 18 demonstrated that multiple, equivalent prognostic signatures for breast cancer can be extracted just by analyzing the same dataset with a different partition in training and test sets, showing the existence of several genes which are practically interchangeable in terms of predictive power. SES along with MMPC are two among the few algorithms, available on hrefhttps://cran.r-project.org/CRAN, that can be used with multiple datasets in a meta-analytic way, following 19.\n\nMXM contains FS algorithms for small sample sized data (MMPC, MMMB, and SES) and for large sample sized data (FBED, gOMP). FBED and gOMP have been adopted for high volume data, going beyond the limits of R. The importance of these customizations can be appreciated by the fact that nowadays large scale datasets are more frequent than before. Since classical FS algorithms cannot handle such data, modifications must be made, in an algorithm level, in a memory efficient manner, in a computer architecture level, and/or in any other way. MXM is using an efficient memory handling R package.\n\nFinally, many utility functions are available, such as constructing a model from the object an algorithm returned, construct a model in general, communication between the input and outputs of the algorithms, long verbose output with useful information, etc. Using hash objects, the computational cost of MMPC and SES is significantly reduced. The univariate associations computed from MMPC, SES and FBED can be interchanged among them and save computational time.\n\nA disadvantage of most MXM’s algorithms is their computational efficiency. Their (algorithmic) order of complexity is comparable to state-of-art FS algorithms, but the nature of the other algorithms is such that many regression models must be fit increasing the computational burden. gOMP, for example, is the most efficient algorithm available in MXM5, because it is residual based and few regression models are fit. However, with clustered/longitudinal data, SES (and MMPC) were shown to scale to tens of thousands and be dramatically faster than LASSO9. Computational efficiency is also programming language-dependent. Most of the algorithms are currently written in R and we are constantly working towards transferring them to C++ so as to decrease the computational cost significantly.\n\nIt is impossible to cover all cases of target variables; we have no algorithms for time series, and do not treat multi-state time-to-event target variables for example, yet we search for R packages that treat other types of target variables and link them to MXM. All algorithms are limited to linear or generalized linear relationships, but we will address this issue in the future. The gOMP algorithm does not accept all types of target variables and works only with continuous predictor variables. This is a limitation of the algorithm, but we plan to address this in the future as well.\n\nCross-validation functions currently exist only for MMPC, SES and gOMP, but performance metrics are not available for all target variables. Left censored data, is an example of target variable whose predictive performance estimation is not offered. A last drawback is that currently MXM does not offer graphical visualization of the algorithms and of the models.\n\nIn terms of sample size, FBED and gOMP are generally advised for large-sample-sized datasets, whereas MMPC and SES are designed mainly for small-sample-sized datasets6. In the case of a large sample size and few features, forward or backward selection are also suggested. In terms of number of features, gOMP is the only algorithm that scales up when the number of features is in the order of the hundreds of thousands. gOMP is also suitable for high volume data that contain a high number of features, really large sample sizes or both. FBED has been customized to handle high volume data as well, but with large sample sizes and only a few thousand features. If the user is interested in discovering more than one set of features, SES is suitable for returning multiple solutions, which are statistically equivalent. With multiple datasets, both MMPC and SES are currently the only two algorithms that can handle some cases (both the target variable and the set of features are continuous). As for the availability of the target variable, MMPC, SES and FBED handle all types of target variables available in MXM, listed in Table 4, while gOMP accepts fewer types of target variables. Regarding the type of features, gOMP currently works with continuous features only, whereas all other algorithms accept both continuous and categorical features. All this information is presented in Table 5.\n\n\nMethods\n\nMXM is an R package that makes use of (depends or imports) many other packages offering regression models\n\n• stats (built-in package): for generalised linear models.\n\n• survival: for survival regression.\n\n• MASS: for negative binomial regression, ordinal regression and MM type regression.\n\n• ordinal: for ordinal regression.\n\n• nnet: for multinomial regression.\n\n• quantreg: for quantile regression.\n\n• lme4: for mixed models.\n\n• geepack: for GEE models.\n\n• coxme: for mixed survival regression models.\n\n• bigmemory: for large volume data.\n\n• doParallel: for parallel computations.\n\n• Rfast: for computational efficiency.\n\nTo help gain computational efficiency, since MXM is not written in C++, MXM imports Rfast21 which was initially created for this purpose. Currently, with little effort, one should be able to plug-in their own regression model into some of the algorithms. We plan to expand this possibility for all algorithms.\n\nMXM contains functions for returning the selected features for a range of hyper-parameters for each algorithm. For example, mmpc.path runs MMPC for multiple combinations of threshold and maxk, and gomp.path runs gOMP for a range of stopping values. The exception is with FBED, for which the user can give a vector of values of K in fbed.reg instead of a single value. Unfortunately, the path of significance levels cannot be determined at a single run.\n\nMMPC and SES have been implemented in such a way that the user has the option to store the results from a single run in a hash object. In subsequent runs, with different hyper-parameters this can lead to significant amounts of computational savings. These two algorithms give the user an extra advantage. They can search for the subset of feature(s) that rendered one more specific feature(s) independent of the target variable by using the function certificate.of.exclusion.\n\nFBED, SES and MMPC are three algorithms sharing some common ground. The list with the results of the univariate associations (test statistic and logged p-value) can be calculated from either algorithm and be passed onto any of them. When one is interested in running many algorithms, this can reduce the computational cost significantly. Note also that the univariate associations in MMPC and SES can be calculated in parallel, with multi-core machines. More FS related functions can be found in MXM’s reference manual and vignettes section available on CRAN.\n\nMXM is distributed as part of the CRAN R package repository and is compatible with Mac OS X, Windows, Solaris and Linux operating systems. Once the package is installed and loaded\n\n\n\nit is ready to be used without internet connection. The system requirements are documented on MXM’s webpage on CRAN.\n\n\nUse cases\n\nWith user-friendliness taken into consideration, extra attention has been put in keeping the functions within the MXM package as consistent as the nature of the algorithms allows for, in terms of syntax, required input objects and parameter arguments. Table 6 contains a list of the current FS algorithms, but we will demonstrate some of them here. In all cases, the arguments \"target\", \"dataset\" and \"test\" refer to the target variable, set of features and type of regression model to be used.\n\nWe will use a variety of target variables and in some examples, we will show the results produced with different regression models. Under no circumstances should the following examples be considered experimental or for the purpose of comparison. They are only for the purpose of algorithms’ demonstration, to give examples of different types of target variables and to show how the algorithms work. All computations took place in a desktop computer with Intel Core i5-4690K CPU @3.50GHz and 32 GB RAM.\n\nThe first dataset we used concerns breast cancer, with 295 women selected from the fresh-frozen–tissue bank of the Netherlands Cancer Institute22. The dataset contains 70 features and the target variable is time to event, with 63 censored values7. We need this information, to be passed as a numerical variable indicating the status (0 = censored, 1 = not censored), for example (1, 1, 0, 1, 1, 1, . . . ). We will make use of the R package survival23 for running the appropriate models (Cox and Weibull regression) and show the FBED algorithm with the default arguments. Part of the output is presented below. Information on the selected features, their test statistic and associated logarithmically transformed p-value, along with some information on the number of regression models fitted is displayed.\n\n\n\nThe above output was produced using Cox regression. If we used Weibull regression instead (test = \"testIndWR\"), the output would be slightly different.\n\n\n\nIn order to avoid small p-values (less than the machine epsilon 10−16) being rounded to 0, their logarithm is computed and returned in the results. This is a crucial and key element of the algorithms because they rely on the correct ordering of the p-values.\n\nThe second dataset we used again concerns breast cancer24 and contains 285 samples over 17,187 gene expressions (features). Since the target variable is binary, logistic regression was employed.\n\n\n\nThe element res presented below is one of the elements of the returned output. The first column shows the selected variables in order of inclusion and the second column is the deviance of each regression model. The first line refers to the regression model with 0 predictor variables (constant term only).\n\n\n\nThe next dataset we will use is NCBI Gene Expression Omnibus accession number GSE910525, which contains 22,283 features about skeletal muscles from 12 normal, healthy glucose-tolerant individuals exposed to acute physiological hyperinsulinemia, measured at 3 distinct time points. Following 9, we will also use SES and not FBED because the sample size is small. The grouping variable, identifying the subject along with the time points are necessary in our case. If the data are repeated measurements or clustered data, i.e. families, where no time is involved, the argument \"reps\" need not be provided. The user has the option to use GLMM26 or GEE27.\n\nThe output of SES (and of MMPC) is long and verbose, but we present the first 10 set of equivalent signatures. The first row is the set of selected features, and every other row is an equivalent set. In this example, the last four columns are the same and only the first changes. This means, that the feature 2683 has 9 statistically equivalent features, (2, 7, ..., 836, ,1117).\n\n\n\nThe next dataset we consider is from Human cerebral organoids recapitulate gene expression programs of fetal neocortex development28. The data are pre-processed RNA-seq, thus continuous data, with 729 samples and 58,037 features. We selected the first feature as the target variable and all the rest were considered to be the features. In this case we used FBED and gOMP, employing the Pearson correlation coefficient because all measurements are continuous.\n\nFBED performed 123, 173 tests and selected 63 features.\n\n\n\ngOMP on the other has was more parsimonious, selecting only 8 features. At this point we must highlight the fact that the selection of a feature was based on the adjusted R2 value. If the increase in the adjusted R2 due to the candidate feature was more than 0.01 or (1/%), the feature was selected.\n\n\n\nThe final example is on discrete valued target variable (count data) for which Poisson and quasi-Poisson regression models will be employed by the gOMP algorithm. The dataset with GEO accession number GSE4777429 contains RNA-seq data with 256 samples and 43,919 features. We selected the first feature to be the target variable and all the rest are the features.\n\nWe ran gOMP using Poisson (test=\"testIndPois\") and quasi Poisson (test=\"testIndQPois\") regression models, but we changed the stopping value to tol=12. Due to over-dispersion (variance > mean), quasi Poisson is appropriate8 because Poisson regression assumes these two quantities are equal. When Poisson was used, 107 features were selected; since the wrong model was used, many false positive features were included, while with the quasi Poisson regression only 10 were selected.\n\n\n\nThe case of ordinal target variable (i.e. very low, low, high, very high) has been treated previously30 for unrevealing interesting features measuring the user perceived quality of experience with YouTube video streaming applications applications and the Quality of Service (target variable) of the underlying network under different network conditions.\n\nMost recently, SES and gOMP were applied in the field of fisheries for identifying the genetic SNP loci that are associated with certain phenotypes of the gilthead seabream (Sparus aurata)31. Measurements from multiple cultured seabream families were taken, thus the data are correlated and GLMM had to be applied. Several of the discovered genes have already been associated with growth in other teleosts or even mice, such as genes MBD5, ACVRIIA and IRF7. The study led to a catalogue of genetic markers that set the ground for understanding growth and other traits of interest in Gilthead seabream, in order to maximize the aquaculture yield.\n\n\nSummary\n\nWe presented the R package MXM and some of its feature selection algorithms. We discussed its advantages and disadvantages and compared it, at a high level, with other competing R packages. We then demonstrated, using real high-dimensional data with a diversity of types of target variables, four FS algorithms, including different regression models in some cases.\n\nThe package is constantly being updated with new functions and improvements being added and algorithms being transferred to C++ to decrease the computational cost. Computational efficiency was mentioned as one of MXM’ disadvantage which we are trying to address. However, computational efficiency is one aspect, and flexibility another. To this end we plan to add of more regression models, more functionalities, options and graphical visualizations.\n\n\nData availability\n\n• The first dataset we used (survival target variable) is available from Computational Cancer Biology.\n\n• The second dataset we used (unmatched case control target variable) is available from GEO.\n\n• The third dataset we used (longitudinal data) is available from GEO.\n\n• The fourth dataset we used (continuous target variable) is available from GEO.\n\n• The fifth dataset we used (count data) is available from GEO.\n\n\nSoftware availability\n\nMXM is available from: https://cran.r-project.org/web/packages/MXM/index.html.\n\nSource code available from: https://github.com/cran/MXM.\n\nArchived source code at time of publication: http://doi.org/10.5281/zenodo.141004332.\n\nLicense: GPL-2.", "appendix": "Grant information\n\nThe research leading to these results has received funding from the European Research Council under the European Union’s Seventh Framework Programme (FP/2007-2013) / ERC Grant Agreement No. 617393.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe would like to acknowledge Stefanos Fafalios, Zacharias Papadovasilakis, Christina Chatzipantsiou, Kleio-Maria Verrou, and Manos Papadakis for their constructive feedback.\n\n\nFootnotes\n\n1In statistics and in the R packages the term \"big data\" is used to refer to such data. In the computer science terminology, big data are of much higher volume and require specific technology. For this reason we chose to use the term \"high volume\" instead of \"big data\".\n\n2MXM is mainly FS oriented, but it offers (Bayesian) network learning algorithms as well. Many feature selection algorithms offered in MXM are Bayesian network inspired.\n\n3We highlight the fact that especially on hrefhttps://cran.r-project.org/CRAN, packages are uploaded at a super-linear rate. Bioconductor is more strict with the addition of new packages. The phenomenon of abandoned or not maintained packages for a long time is not at all unusual. Such an example is \"biospear\", removed from CRAN (archived) in the 30th of April 2018. On the other hand we added a package that performs FS without mentioning this in its title.\n\n5For this long list of available target variables and regression models, expanding Table 4, see Guide on performing FS with the R package MXM.\n\n5In our anecdotal experiments it has superseded the LASSO implementation in the package glmnet20 in both time and performance.\n\n6To the best of our knowledge there are not many FS algorithms dealing with small sample sized data.\n\n7Censoring occurs when partial information about some observations is available. It might be the case that some individuals will experience the event after completion of the study. Or when an individual is not part of the study for anymore, for a reason other than the occurrence of the event of interest. In a study about cancer, for example, some patients may die of another cause, e.g. another disease or car accident for example. The survival times of those patients has been recorded, but offer limited information.\n\n8Negative binomial regression, test=\"testIndNB\" is another alternative option.\n\n\nReferences\n\nTsamardinos I, Aliferis CF: Towards principled feature selection: relevancy, filters and wrappers. In AISTATS. 2003. Reference Source\n\nLagani V, Athineou G, Farcomeni A, et al.: Feature Selection with the R Package MXM: Discovering Statistically-Equivalent Feature Subsets. J Stat Softw. 2017; 80(7). Publisher Full Text\n\nTsamardinos I, Aliferis CF, Statnikov AR: Algorithms for Large Scale Markov Blanket Discovery. In FLAIRS Conference. 2003. Reference Source\n\nTsamardinos I, Aliferis CF, Statnikov A: Time and sample efficient discovery of Markov Blankets and direct causal relations. In Proceedings of the ninth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM. 2003; 673–678. Publisher Full Text\n\nAliferis CF, Statnikov A, samardinos I, et al.: Local causal and Markov Blanket induction for causal discovery and feature selection for classification part II: Analysis and extensions. J Mach Learn Res. 2010; 11: 235–284. Reference Source\n\nTsamardinos I, Brown LE, Aliferis CF: The Max-Min Hill-Climbing Bayesian network structure learning algorithm. Mach Learn. 2006; 65(1): 31–78. Publisher Full Text\n\nLagani V, Tsamardinos I: Structure-based variable selection for survival data. Bioinformatics. 2010; 26(15): 1887–1894. PubMed Abstract \| Publisher Full Text\n\nLagani V, Kortas G, Tsamardinos I: Biomarker signature identification in \"omics\" data with multi-class outcome. Comput Struct Biotechnol J. 2013; 6(7): e201303004. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTsagris M, Lagani V, Tsamardinos I: Feature selection for high-dimensional temporal data. BMC Bioinformatics. 2018; 19(1): 17. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGroll A, Tutz G: Variable selection for generalized linear mixed models by L1-penalized estimation. Stat Comput. 2014; 24(2): 137–154. Publisher Full Text\n\nBorboudakis G, Tsamardinos I: Forward-backward selection with early dropping. arXiv preprint arXiv: 1705.10770. 2017. Reference Source\n\nChen S, Billings SA, Luo W: Orthogonal least squares methods and their application to non-linear system identification. Int J Control. 1989; 50(5): 1873–1896. Publisher Full Text\n\nPati YC, Rezaiifar R, Krishnaprasad PS: Orthogonal matching pursuit: Recursive function approximation with applications to wavelet decomposition. In Signals, Systems and Computers, 1993. 1993 Conference Record of The Twenty-Seventh Asilomar Conference on. IEEE. 1993; 40–44. Publisher Full Text\n\nDavis G: Adaptive nonlinear approximations. PhD thesis, New York University, Graduate School of Arts and Science, 1994. Reference Source\n\nSchwarz G: Estimating the dimension of a model. Ann Stat. 1978; 6(2): 461–464. Publisher Full Text\n\nChen J, Chen Z: Extended bayesian information criteria for model selection with large model spaces. Biometrika. 2008; 95(3): 759–771. Publisher Full Text\n\nTsamardinos I, Lagani V, Pappas D: Discovering multiple, equivalent biomarker signatures. In Proceedings of the 7th conference of the Hellenic Society for Computational Biology Bioinformatics, Heraklion, Crete, Greece. 2012. Reference Source\n\nEin-Dor L, Kela I, Getz G, et al.: Outcome signature genes in breast cancer: is there a unique set? Bioinformatics. 2005; 21(2): 171–178. PubMed Abstract \| Publisher Full Text\n\nLagani V, Karozou AD, Gomez-Cabrero D, et al.: A comparative evaluation of data-merging and meta-analysis methods for reconstructing gene-gene interactions. BMC Bioinformatics. 2016; 17 Suppl 5: 194. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFriedman J, Hastie T, Tibshirani R: Regularization Paths for Generalized Linear Models via Coordinate Descent. J Stat Softw. 2010; 33(1): 1–22. PubMed Abstract \| Free Full Text\n\nPapadakis M, Tsagris M, Dimitriadis M, et al.: Rfast: A Collection of Efficient and Extremely Fast R Functions. R package version 1.9.0. 2018. Reference Source\n\nvan de Vijver MJ, He YD, van’t Veer LJ, et al.: A gene-expression signature as a predictor of survival in breast cancer. N Engl J Med. 2002; 347(25): 1999–2009. PubMed Abstract \| Publisher Full Text\n\nTherneau TM: A Package for Survival Analysis in R. version 2.42-6. 2018. Reference Source\n\nWang Y, Klijn JG, Zhang Y, et al.: Gene-expression profiles to predict distant metastasis of lymph-node-negative primary breast cancer. Lancet. 2005; 365(9460): 671–679. PubMed Abstract\n\nColetta DK, Balas B, Chavez AO, et al.: Effect of acute physiological hyperinsulinemia on gene expression in human skeletal muscle in vivo. Am J Physiol Endocrinol Metab. 2008; 294(5): E910–E917. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBates D, Mächler M, Bolker B, et al.: Fitting linear mixed-effects models using lme4. arXiv preprint arXiv: 1406.5823. 2014. Reference Source\n\nHøjsgaard S, Halekoh U, Yan J: Package geepack. R package version 1.2-0. 2015. Reference Source\n\nCamp JG, Badsha F, Florio M, et al.: Human cerebral organoids recapitulate gene expression programs of fetal neocortex development. Proc Natl Acad Sci U S A. 2015; 112(51): 15672–15677. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSEQC/MAQC-III Consortium: A comprehensive assessment of RNA-seq accuracy, reproducibility and information content by the Sequencing Quality Control Consortium. Nat Biotechnol. 2014; 32(9): 903–14. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKatsarakis M, Teixeira RC, Papadopouli M, et al.: Towards a causal analysis of video qoe from network and application qos. In Proceedings of the 2016 workshop on QoE-based Analysis and Management of Data Communication Networks. ACM. 2016; 31–36. Reference Source\n\nKyriakis D, Kanterakis A, Manousaki T, et al.: Scanning of genetic variants and genetic mapping of phenotypic traits in gilthead seabream (sparus aurata). In preparation. 2018.\n\nTsagris M, Tsamardinos I, Lagani V, et al.: Feature Selection (Including Multiple Solutions) and Bayesian Networks (Version 1.3.9). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1410043" }	[ { "id": "38571", "date": "08 Oct 2018", "name": "Thodoris Kypraios", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nSummary\nThe paper is concerned with the method of feature selection using the R package MXM. The package appears to be fairly versatile in the sense that it can handle a huge variety of types of data. It can be very useful for applied researchers and, at the very least, it is another tool in the toolbox for the applied statistician.\n\nI believe that the paper it will be a useful addition in the literature. However, I found difficult to read/understand in places. For example, is the MXM a package that only includes methodology that has been developed by the authors or does it include methods that have been developed by other researchers too? Either way is fine, but it would be helpful to clarify.\nAnother major issue for me is that the paper is flooded with acronyms that I believe most users will be unfamiliar with. It would improve the paper's presentation significantly, if there is some explanation (a couple of sentences per method would suffice) about each of them.\n\nBelow are some more specific points to consider:\n\nAbstract\n\nIt it not clear when one first reads what \"b) it contains a variety of regression models to plug into the feature selection algorithms;\" means. c), \"equivalent\" in what sense? \"it includes memory efficient algorithms for high volume data, data that cannot be loaded into R\" -> \"it includes memory efficient algorithms for high volume data that often cannot be loaded easily into R\"?\n\nIntroduction\n\"and easier to understand and interpret;\" -> \"and often easier to understand and interpret;\" \"small portion\" -> \"small proportion\"? It would be good to Table 1 to have the totals. \"These algorithms have been tested and compared with other state-of-the-art algorithms under different scenarios and types of data.\" Any references? I find hard to read the \"Comparisons of MXM’ FS algorithms with other FS algorithms\" because there are a bunch of acronyms used that I (and presumably other readers) don't know what they mean; it would be good to say a few words about what each algorithm is doing. \"anecdotal\" -> preliminary? It would be good to say what is g-OMP (as well as the other methods) doing?\n\nMethods\nIn the example \"Survival (or time-to-event) target variable\", a survival model is fitted and the MXM package is used and its output is presented, but I am unclear as to what is the objective (in terms of the data analysis) and what does the output mean. The above comments applies to the above datasets; it is crucial that reader knows what is the statistical objective first, and then to explain what the outcome of the package means. page 11: \"gOMP on the other\" -> \"gOMP on the other _hand_\" ?\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4747", "date": "30 Sep 2019", "name": "Michail Tsagris", "role": "Author Response", "response": "We are grateful to the reviewers for their on-the-spot comments.The paper is concerned with the method of feature selection using the R package MXM. The package appears to be fairly versatile in the sense that it can handle a huge variety of types of data. It can be very useful for applied researchers and, at the very least, it is another tool in the toolbox for the applied statistician. Comment: I believe that the paper it will be a useful addition in the literature. However, I found difﬁcult to read/understand in places. For example, is the MXM a package that only includes methodology that has been developed by the authors or does it include methods that have been developed by other researchers too? Either way is ﬁne, but it would be helpful to clarify. Reply: We have added an extra column in Table 5 giving this information. Also, in Section “The MXM’s FS algorithms and comparison with other FS algorithms” where we describe the algorithms we have added the relevant references. Comment: Another major issue for me is that the paper is ﬂooded with acronyms that I believe most users will be unfamiliar with. It would improve the paper’s presentation signiﬁcantly, if there is some explanation (a couple of sentences per method would sufﬁce) about each of them. Reply: We have added the explanation of all acronyms when the algorithm is ﬁrst mentioned. Also in Section “The MXM’s FS algorithms and comparison with other FS algorithms” we brieﬂy mention how they work. Below are some more speciﬁc points to consider:Abstract Comment: It it not clear when one ﬁrst reads what \"b)it contains a variety of regression models to plug into the feature selection algorithms;\" means. Reply: We have added a sentence inside parentheses giving some examples of target variables and appropriate regression models explaining this sentence. Comment: c), \"equivalent\" in what sense? Reply: We added a sentence inside parentheses explaining this sentence. Equivalent refers to the“information” a feature contains. If for example the information gain of a variable is not signiﬁcant it could be attributed to another feature that contains the same information as this one. Think for example collinear variables, such as length of left hand and length of the right hand. Either hand can be used in a regression model. This phenomenon is prevalent, but also not highly examined, in bioinformatics where the selected genes may not be the ones expected by the biologists, only because some equivalent genes were selected. Only few feature selection algorithms return all equivalent features and SES is one of them. Comment: \"it includes memory efﬁcient algorithms for high volume data, data that cannot be loaded into R\" -> \"it includes memory efﬁcient algorithms for high volume data that often cannot be loaded easily into R\"? Reply: We have added a sentence inside parentheses explaining this sentence. The term “big data” has wrongfully been used in many instances to denote large scale data. What we mean by high volume data is data whose size in Gb equals or exceeds the size of the available RAM in one’s computer and hence cannot be loaded into R. Introduction •Comment: \"and easier to understand and interpret;\" -> \"and often easier to understand and interpret;\" • Reply: We thank the reviewer for this grammatical suggestion. We changed it. • Comment: \"small portion\" -> \"small proportion\"? Reply: We did not change this word. We consider the word “portion” as a synonym for “share”. The word “proportion” is rather further away from our interpretation and the meaning of the word we wanted to use. Comment: It would be good to Table 1 to have the totals. Reply: We added the total (184 packages) in the caption of Table 1. Comment: \"These algorithms have been tested and compared with other state-of-the-art algorithms under different scenarios and types of data.\" Any references? Reply: For each algorithm separately, we have added the relevant reference. Comment: I ﬁnd hard to read the \"Comparisons of MXM’ FS algorithms with other FS algorithms\" because there are a bunch of acronyms used that I (and presumably other readers) don’t know what they mean; it would be good to say a few words about what each algorithm is doing. Reply: We thank the reviewer for this comment. We have expanded this section by adding a short description for each algorithm. Comment: \"anecdotal\" -> preliminary? Reply: We changed the whole sentence and refer to the relevant paper that is a draft paper available on bioRxiv. Comment: It would be good to say what is g-OMP (as well as the other methods) doing? Reply: We have added a short description of what each feature selection algorithm does. Methods Comment: In the example \"Survival (or time-to-event) target variable\", a survival model is ﬁtted and the MXM package is used and its output is presented, but I am unclear as to what is the objective (in terms of the data analysis) and what does the output mean. Reply: We thank the reviewer for this important comment. We have modiﬁed this example and the others. In addition, we have added a paragraph in the beginning of this section explaining the goal of this section. Comment: The above comments applies to the above datasets; it is crucial that reader knows what is the statistical objective ﬁrst, and then to explain what the outcome of the package means. Reply: Please see previous comment. Comment: page 11: \"gOMP on the other\" -> \"gOMP on the other hand\"? Reply: We added the word “hand”." } ] }, { "id": "42926", "date": "28 Jan 2019", "name": "Huitong Qiu", "expertise": [ "Reviewer Expertise statistical machine learning", "feature selection", "high dimensional data", "graphical models", "time series analysis", "clinical trial design" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript introduces a new R package, MXM, that offers a variety of feature selection algorithms in regression models. The new package presents relevant contribution to the toolbox of feature selection algorithms by covering more types of target variables than most existing packages, accommodating data with small or big sample sizes, and providing additional functionalities and utility features. The manuscript also demonstrates the usage of several functions in the package by analyzing real data.\nRegarding the presentation of the manuscript, I share the same concern with reviewer Thodoris that the acronyms in the paper come with poor explanation. Although several of the acronyms are explained in Table 4 and Table 6, most of appearances of these acronyms have no reference to these tables or explanations. This makes the paper unnecessarily hard to read.\nSecondly, I think the demonstration of the usage of the package could be more informative with more interpretations. For example, how can we interpret the p-values, adjusted R-squares, and deviance in the results of the model fittings? High level descriptions of the algorithms behind the demonstrated functions could also help the reader better understand the results.\nBelow are comments to specific places in the manuscript:\nThe 3rd paragraph in Introduction mentions that statistical tests (likelihood ratio, Wald, permutation based) can be plugged into the feature selection algorithms that work with small sample sized data. The aforementioned tests traditionally rely on large sample theory. I wonder what adjustments are needed to accommodate small sample sizes? Tables 1, 2, and Figure 1 shows statistics of target variable types supported by a number of R packages. What algorithm was used to identify the types of target variables accepted by an R package? What does “multiple datasets” mean as in “Only 2 (1.08%) R packages teat the case of FS with multiple datasets ...” (last sentence of Paragraph 1 on Page 5)? I find the last paragraph on Page 5 particularly uneasy to read due to the acronyms used without explanation. Some of these are later explained in Table 6 while a couple are not (e.g., MMHC). I think it would be helpful to make a comprehensive table explaining the algorithm acronyms used in this paper, and refer to the table here. In the last paragraph on Page 5, there are multiple places that compare algorithms in terms of “(predictive) performance”. I think it would be more informative to clarify what performance metrics we are looking at. In Paragraph 5 on Page 6, the manuscript states that “SES ... returns multiple (statistically equivalent) sets of predictor variables ...”. What does statistically equivalent mean? What does the package name, MXM, stand for? The second last paragraph on Page 6 is a bit confusing to me. The paragraph starts with pointing out computational efficiency as a disadvantage of MXM. However, this is followed by explaining why gOMP is efficient, and pointing out that SES and MMPC scales better and run faster than LASSO package. These seem like advantages in computational efficiency. In Paragraph 2 of “FS-related functions” section: for MMPC and SES, storing the results from one run and passing it to subsequent runs can lead to significant computational savings. Are there savings because the trained models serve as good starting points in the subsequent runs? On Page 9 in the output of MXM::fbed.reg, there are p-values associated with each selected variable. How can we interpret these p-values? It seems that these p-values are not adjusted for the extra degrees of freedom from feature selection.\nBelow are some minor suggestions/typo fixes:\nIn Abstract – “The R package MXM is such an example, which offers ...”: It’s unclear to me what “such an example” refers to. It’s clearer to state “The R package MXM offers ...” directly. Second paragraph in Introduction: “For example, packages that accept few or specific types of target variables.” → “For example, some packages accept few or specific types of target variables.” (make it a complete sentence.) In “MXM versus other R packages” section: “... can treat at least one type of target variable, ...” → “... can treat at least k types of target variables, for k = 1, 2, ..., 8, ...” In the last paragraph on Page 5, does “BN” refer to Bayesian network (which also appears in the same paragraph)? In Paragraph 5 on Page 6 – “... making it one one of the few FS algorithms ...”: remove the extra “one”. In Paragraph 6 on Page6 – “MXM is using an efficient memory handling R package.”: please cite the package. Last Paragraph on Page 7: “generalised” → “generalized” to be consistent with other places in the paper. First paragraph in Section “Use cases”: the argument “test” refers to the type of regression model to be used. Why not name the argument something like “model type”? Some of the citations are not easily distinguishable from footnotes. e.g., citation 21 for Rfast on Page 8 and footnote 7 are both superscripts. Paragraph 2 on Page 11: “gOMP on the other has ...” → “gOMP on the other hand ...”. First paragraph on Page 12 – “YouTube video streaming applications applications”: duplicate “applications”.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4748", "date": "30 Sep 2019", "name": "Michail Tsagris", "role": "Author Response", "response": "We are grateful to the reviewers for their on-the-spot comments which we have addressed. The manuscript introduces a new R package, MXM, that offers a variety of feature selection algorithms in regression models. The new package presents relevant contribution to the toolbox of feature selection algorithms by covering more types of target variables than most existing packages, accommodating data with small or big sample sizes, and providing additional functionalities and utility features. The manuscript also demonstrates the usage of several functions in the package by analyzing real data. Comment: Regarding the presentation of the manuscript, I share the same concern with reviewer Thodoris that the acronyms in the paper come with poor explanation. Although several of the acronyms are explained in Table 4 and Table 6, most of appearances of these acronyms have no reference to these tables or explanations. This makes the paper unnecessarily hard to read. Reply: We have added the interpretation of the acronyms when they ﬁrst appear, at various places within the text. Comment: Secondly, I think the demonstration of the usage of the package could be more informative with more interpretations. For example, how can we interpret the p-values, adjusted R-squares, and deviance in the results of the model ﬁttings? High level descriptions of the algorithms behind the demonstrated functions could also help the reader better understand the results. Reply: We thank the reviewer for this comment. We have added a small description of each algorithm does in Section “The MXM’s FS algorithms and comparison with other FS algorithms”. This was also highlighted by Prof Kypraios. Also, in Section “Advantages and disadvantages of MXM’s FS algorithms” we have added a small paragraph regarding the p-values produced by the algorithms. Below are comments to speciﬁc places in the manuscript: Comment: The 3rd paragraph in Introduction mentions that statistical tests(likelihood ratio, Wald, permutation based) can be plugged into the feature selection algorithms that work with small sample sized data. The aforementioned tests traditionally rely on large sample theory. I wonder what adjustments are needed to accommodate small sample sizes? Reply: We have added a foot note in page 5 regarding this. There as on we added this information at this point was because we discuss the suitability of the algorithms based on the sample size. For example, we say the MMPC is more suitable for small sample sized data and explain why. FBED on the other hand is devised for large sample sizes. Comment: Tables 1, 2, and Figure 1 shows statistics of target variable types supported by a number of R packages. What algorithm was used to identify the types of target variables accepted by an R package? Reply: We did this manually. We searched on CRAN and went through all packages one by one. We did this process twice to make sure we did not omit any package. Bear in mind that this search was made a few months ago. Comment: What does “multiple datasets” mean as in “Only 2 (1.08%) R packages teat the case of FS with multiple datasets ...” (last sentence of Paragraph 1 on Page 5)? Reply: We added a footnote at this point explaining what we mean by multiple datasets and how we perform feature selection in this case. The 1.08% means that only 2 R packages (available on CRAN or Bioconductor) perform feature selection with multiple datasets. Comment: I ﬁnd the last paragraph on Page 5 particularly uneasy to read due to the acronyms used without explanation. Some of these are later explained in Table 6 while a couple are not (e.g., MMHC). I think it would be helpful to make a comprehensive table explaining the algorithm acronyms used in this paper, and refer to the table here. Reply: We thank the reviewer for this comment. We have explained all acronyms in various place, whenever the relevant algorithm is ﬁrst mentioned. Comment: In the last paragraph on Page 5, there are multiple places that compare algorithms in terms of “(predictive)performance”. I think it would be more informative to clarify what performance metrics we are looking at. Reply: We have added a couple a sentence mentioning some predictive performance metrics for some target variables. Comment: In Paragraph 5 on Page 6, the manuscript states that “SES ... returns multiple (statistically equivalent) sets of predictor variables ...”. What does statistically equivalent mean? Reply: This comment was raised by the other reviewer as well. In the Abstract we have added a sentence inside parentheses brieﬂy explain this. Also in Section “The MXM’s FS algorithms and comparison with other FS algorithms” we have added one sentence when mention the SES algorithm (this is in magenta colour). This comment was also made by the ﬁrst reviewer and we have answered this previously. Comment: What does the package name, MXM, stand for? Reply: We thank the reviewer for this comment. We added the meaning of this acronym in the 4rth footnote of page 2. It stands for Mens ex Machina. Comment: The second last paragraph on Page 6 is a bit confusing to me. The paragraph starts with pointing out computational efﬁciency as a disadvantage of MXM. However, this is followed by explaining why gOMP is efﬁcient, and pointing out that SES and MMPC scales better and run faster than LASSO package. These seem like advantages in computational efﬁciency. Reply: We thank the reviewer for this clariﬁcation. Indeed the message is diluted. We have reworded this point, towards the end of page 5. Comment: In Paragraph 2 of “FS-related functions” section: for MMPC and SES, storing the results from one run and passing it to subsequent runs can lead to signiﬁcant computational savings. Are there savings because the trained models serve as good starting points in the subsequent runs? Reply: We have added a few sentences clarifying this point. All p-values are stored to avoid implementing all tests again in subsequent runs. Comment: On Page 9 in the output of MXM::fbed.reg, there are p-values associated with each selected variable. How can we interpret these p-values? It seems that these p-values are not adjusted for the extra degrees of freedom from feature selection. Reply: We do recognize this problem and are aware that the distribution of the test statistics is not the assumed one [1] and their associated p-values can be small [2]. We have added a small discussion on the p-values at the bottom of page 4-top of page 5 regarding this. The p-values are not FDR corrected, but we refer to the papers that mention this issue as well. With MMPC for example, there is no need to apply FDR because the algorithm controls the FDR [3]. [4] discusses methods that address this issue and they conclude that the FBED algorithm is orthogonal to those methods and could be used in conjunction with them. Closing this issue we will mention that it is not easy to control FDR in the context of feature selection; it is an open problem. It is commonly accepted in the statistical literature, and not only, that this leads to regression coefﬁcients are overestimated. A solution would be to ﬁt a LASSO regularised model, for example, and tune the penalty algorithm via cross-validation. However, the problem of FDR is not totally addressed, but rather bypassed. Below are some minor suggestions/typo ﬁxes: Comment: In Abstract – “The R package MXM is such an example, which offers ...”: It’s unclear to me what “such an example” refers to. It’s clearer to state “The R package MXM offers ...” directly. Reply: We thank the reviewer for this suggestion. We have reworded this sentence. Comment: Second paragraph in Introduction: “For example, packages that accept few or speciﬁc types of target variables.” -> “For example, some packages accept few or speciﬁc types of target variables.” (make it a complete sentence.) Reply: We thank the reviewer for this syntactical mistake which we have now corrected. Comment: In “MXM versus other R packages” section: “... can treat at least one type of target variable, ...” -> “... can treat at least k types of target variables, for k = 1, 2, ..., 8, ...” Reply: We have changed this sentence as requested. Comment: In the last paragraph on Page 5, does “BN” refer to Bayesian network (which also appears in the same paragraph)? Reply: We substituted the BN to Bayesian network. Comment: In Paragraph 5 on Page 6 – “... making it one one of the few FS algorithms ...”: remove the extra “one”. Reply: We have removed the extra “one”. Comment: In Paragraph 6 on Page6 – “MXM is using an efﬁcient memory handling R package.”: please cite the package. Reply: We have mentioned and cited the relevant package. Comment: Last Paragraph on Page 7: “generalised” -> “generalized” to be consistent with other places in the paper. Reply: We changed everything to “generalised”. Comment: First paragraph in Section “Use cases”: the argument “test” refers to the type of regression model to be used. Why not name the argument something like “model type”? Reply: We have reworded this part of the manuscript and added some more information about its target. Comment: Some of the citations are not easily distinguishable from footnotes. e.g., citation 21 for Rfast on Page 8 and footnote 7 are both superscripts. Reply: We did not spot this issue in the paper. # Comment: Paragraph 2 on Page 11: “gOMP on the other has ...” -> “gOMP on the other hand ...”. Reply: We changed this as requested by the ﬁrst reviewer also. Comment: First paragraph on Page 12 – “YouTubevideostreamingapplicationsapplications”: duplicate “applications”. Reply: We removed the duplicate. [1] Hastie Trevor, Tibshirani Robert and Friedman JH. The elements of statistical learning: data mining, inference, and prediction. New York: Springer, 2009.[2] Frank E Harrell Jr. Regression modeling strategies. Springer. 2017.[3] Ioannis Tsamardinos and Laura E Brown.Bounding the False Discovery Rate in Local Bayesian Network Learning. In AAAI, pages 110-1105, 2008.[4] Giorgos Borboudakis and Ioannis Tsamardinos. Forward-backward selection with early dropping. Journal of Machine Learning Research, 20(8):1–39, 2019." } ] } ]	1	https://f1000research.com/articles/7-1505
https://f1000research.com/articles/8-1699/v1	27 Sep 19	{ "type": "Research Article", "title": "Streptococcus pneumoniae serotype specific anti-microbial susceptibility profiles among PCV-10 vaccinated and unvaccinated children attending Gertrude’s Children’s Hospital: a cross-sectional study", "authors": [ "Michael Walekhwa", "Margaret Muturi", "Eucharia Kenya", "Beatrice Kabera", "Margaret Muturi", "Eucharia Kenya", "Beatrice Kabera" ], "abstract": "Background: The spread of antimicrobial resistance threatens effective control and treatment of pneumococcal disease worldwide. In Kenya, an estimated one in every five children dies from pneumococcal disease every year. Of these, ≥50% are attributable to antibiotic resistance. Consequently, the WHO has recommended that continuous regional surveillance be done to detect early resistance to available antibiotics and make necessary changes. We therefore investigated antimicrobial susceptibility patterns of Streptococcus pneumoniae among PCV-10 vaccinated and unvaccinated children ≤5 years old at Gertrude's Children’s Hospital. Methods: A 0.5 McFarland standard of freshly subcultured organisms were inoculated on Mueller–Hinton plates with 5% sheep blood agar. A standard disk dispenser was used to dispense various antibiotic disks on the Mueller–Hinton agar plate. Incubation was done overnight (20-24 hours) at 37oC in 5% CO2 and clearance zones read using a Vanier caliber. Antimicrobials tested included vancomycin (30µg, ≥17mm); erythromycin (15µg, ≥21mm); clindamycin (2µg, ≥19mm); oxacillin (1µg, ≥19mm) and ceftriaxone (1µg, ≥30mm). Results: Thirty nine (92.86%) Streptococcus pneumoniae isolates were susceptible to erythromycin; 39 (92.86%) were susceptible to vancomycin; eight (19.86%) Streptococcus pneumoniae isolates were susceptible to oxacillin, while 34 (80.95%) were non-susceptible; 40 (95.24%) isolates were susceptible to clindamycin; and 24 (57.86%) isolates were susceptible to ceftriaxone, while 18 (42.86%) were non-susceptible. Children who attended daycare centers exhibited a four-fold significant risk of being resistant to ceftriaxone. All antibiotics studied were effective against Streptococcus pneumoniae except oxacillin and ceftriaxone, which exhibited high levels of non-susceptibility. Attendance of daycare centers, consumption of antibiotics two weeks prior to collection of sample and subject age were shown to be associated with an increased risk of Streptococcus pneumoniae being resistant to penicillins and ceftriaxone. Conclusions: The law guiding use of antibiotics in Kenya should be meritoriously enforced to curb abuse of the available antibiotics.", "keywords": [ "Streptococcus pneumoniae", "Antimicrobial Resistance", "Optochin", "Kirby Bauer Test" ], "content": "Introduction\n\nGlobally, there is a growing concern over the burden of antimicrobial resistance (AMR) because it is presently estimated to account for more than 700,000 deaths per year worldwide (WHO, 2014). Hypothetically, studies have reported that if no appropriate measures are taken to arrest this unfortunate trend, AMR may cost up to 10 million lives and around US $100 trillion per year by 2050 (de Kraker et al., 2016). What is more appalling is that, contrary to some other health trepidations, AMR is a problem that concerns every country regardless of its level of income and development as resistant pathogens do not respect borders and jurisdictions (Cole, 2016).\n\nDiverse forms of pneumococcal disease are rated among the highest vaccine preventable etiologies of child morbidity and mortality globally (Lowth, 2015). This is true despite availability of effective vaccines and antibiotic agents; it is therefore a matter of grave concern that Streptococcus pneumoniae is developing resistance to available antimicrobial agents. Paucity of quality and verifiable data on the pneumococci antimicrobial susceptibility profiles is a precursor for inadequate treatment guidelines, especially for countries that are still largely developing (Assamala, 2014). The gap in public health capacity is also an issue, given the changing resistance machineries and the emergence of multidrug-resistant Streptococcus pneumoniae that can only be detected through systematic screening in quality assured microbiology laboratories (O’Neill, 2016).\n\nOvertime, there has been unswerving universal increase in antibiotic resistance by Streptococcus pneumoniae isolates, a trend that has made treatment options more challenging (Henriques-Normark & Tuomanen, 2013). The bacterium’s polysaccharide capsule is a major virulence factor (Mitchell & Mitchell, 2010). The bacterial serotype is determined on the basis of the capsule's antigenic features, which have also been expansively correlated with AMR patterns. Although Streptococcus pneumoniae antimicrobial resistance is a wide-reaching concern, low and middle income countries (LMICSs) are vulnerable to exceptional difficulties because substitute and/or effective therapeutic choices are either unattainable or too expensive for them to afford (Reardon, 2014). Experiential prescription of antibiotics for treatment of various forms of pneumococcal disease without data on antimicrobial susceptibility profiles is a very common practice, especially in Africa (Feldman & Anderson, 2016). Purchase of antibiotics over the counter and failing to comply with rules of antibiotic consumption and/or use significantly augments occurrence of AMR in Kenya (Donkor & Badoe, 2014). Unregulated and illicit hawking of major antibiotics at bus-stops by unlicensed persons and back-street manufacturing of counterfeit antimicrobial agents contribute majorly to pneumococci AMR (Laxminarayan et al., 2013).\n\nThe end result of ‘abuse’ of antimicrobial agents meant for treatment of Streptococcus pneumoniae infections is apparent; the pneumococcus has connivingly developed resistance to novel antibiotic agents (Ventola, 2015). Since the rates of Streptococcus pneumoniae AMR vary on the basis of geographic location, time, age, and site of infection, AMR surveillance must be incessant and guidelines be derived from local AMR data (Feldman & Anderson, 2016). As AMR impinges on the quality of patient care, concerted efforts must be made to understand the best way to circumvent it. To treat various forms of pneumococcal disease, antibiotics are the most commonly prescribed drugs in hospitals. Judicious use of antimicrobial agents is therefore one of the major avenues that can be employed to arrest the worrying trend of AMR.\n\nAn audit of Streptococcus pneumoniae antibiotic susceptibility patterns is imperative in the execution of a lucid and pragmatic antibiotic strategy. We therefore sought to analyze antimicrobial resistance profiles to commonly used antibiotics meant for treatment of pneumococcal disease among children ≤5 years in Nairobi County.\n\n\nMethods\n\nThis was a descriptive cross-sectional study. Streptococcus pneumoniae anti-microbial susceptibility patterns among PCV-10 vaccinated and unvaccinated children below five years of age were measured between May 2017 to February 2018.\n\nEthical approval for this study was given by the Kenyatta University Ethics Review Committee {KU/ERC/APPROVAL/VOLUME.1 (12)} and permission to carry out the study obtained from the National Commission of Science Technology and Innovation (NACOSTI/P/17/65428/15801). Gertrude's Children’s Hospital (GCH) research committee issued permission for the study to be conducted at the GCH outpatient clinic (GCH/ERB/VOLMMXVII/121). Biological mothers and/or legal guardians to eligible study subjects were approached to sign a written informed consent to allow their children participate in the study; this was done after reading and having the study explained to them. Participation in the study was completely on voluntary terms.\n\nThis study was conducted among children attending GCH, Nairobi County. GCH is the leading stand-alone hospital in East and Central Africa specializing in pediatric care. The hospital is accredited by the Joint Commission International (JCA). Streptococcus pneumoniae laboratory isolation, disk diffusion for AMR profiling and isolate stocking was done at the GCH main laboratory, while Streptococcus pneumoniae serotyping was done at KEMRI Wellcome Trust, Kilifi.\n\nTo take part in this study, biological mothers and/or legal guardians were approached at the GCH outpatient clinic when they brought their children for medical attention. For inclusion in this study, the subject was required to: be at most five years of age; be a resident of either Nairobi or Kiambu Counties; have been vaccinated with either three doses of PCV-10 or none (any child who had received less than three doses was considered ineligible); not be consuming any antibiotics at the time of the study; be clinically diagnosed by a physician at the site as having some form of pneumococcal disease; and the parent or guardian needed to voluntarily sign an informed consent form. Subjects who had known HIV/AIDS infection or had any immunosuppressive conditions were excluded from the study.\n\nTo determine the minimum sample size, the Fisher’s formula (1998) was used, with a prevalence rate of 16% (Tuti et al., 2017) being the current occurrence rate of pneumococcal disease in Kenya.\n\n\n\nWhere: n = desired minimal sample size; z = standard normal deviation = 1.96 (from the tailed normal table); p^ = prevalence rate; and m = desired degree of accuracy at 95% confidence level = 0.05. n = 1.962 x 0.16 (1-0.16) / 0.052 = 206; therefore, n = 206.\n\nA standardized questionnaire was used to collect data on the subjects’ sociodemographic features. The characteristics assessed included whether the child attended daycare or not, exposure to both cigarette and cooking smoke, breastfeeding frequency during the first six months after birth (classified as: none, meaning the child was never breastfed at all; moderate, meaning the child partially breastfed and partially given other food; and exclusive, meaning the child was entirely dependent on breast-milk), child’s age and PCV-10 immunization status.\n\nNasopharyngeal swabs collected using flocked cotton swabs (Copan) were the specimen of choice for this study. All swabs were initially suspended in Amies Media and transported to the laboratory within three hours of collection. Each swab was inoculated onto a selective gentamicin blood agar (GBA) enriched with 5% sheep blood. The plates were incubated at 37°C in 5% CO2 and examined at 16–24 hours and then again at 40–48 hours for growth of Streptococcus pneumoniae.\n\nIsolates were identified as Streptococcus pneumoniae on the basis of colony morphology (mucoid, draughtsman's appearance, α-haemolysis), susceptibility to optochin (ethyl hydrocupreine hydrochloride) and, where necessary, solubility in bile salts (desoxycholate salts). The isolation of a single colony indicated nasopharyngeal carriage.\n\nAll colonies on GBA plates demonstrating Streptococcus pneumoniae characteristics were evenly inoculated on plain blood agar (BA) plates. This was done using a sterile inoculating loop. An optochin disk was aseptically placed on the preparation using a sterile antibiotic dispenser and incubated for 24–48 hours at 37°C in 5% CO2 atmosphere. After 24–48 hours, a clearance zone of ≥14mm around the optochin disk was a confirmation that the organism was Streptococcus pneumoniae. Plates with clearance zones <14mm but ≥6mm were further subjected to a bile solubility test. Bile soluble stocks confirmed that the organism was Streptococcus pneumoniae.\n\nUsing a sterile cotton inoculating swab, Streptococcus pneumoniae colonies from the blood agar plate were added to 1.0ml of 0.85% saline solution to achieve 1.0 McFarland standard turbidity. The cell suspension was divided into two equal volumes of 0.5ml/tube. Then, 0.5 ml of 2% sodium desoxycholate solution (bile salts) were added to one tube and 0.5 ml of 0.85% saline solution to another tube. The two tubes were gently mixed and incubated for up to two hours at 37°C in 5% CO2. The tubes were vortexed while observing for any clearance of turbidity within 10 minutes and subsequently checked until two hours later. Any clearance in the tube containing bile salts and not in the saline tube confirmed that the organism was Streptococcus pneumoniae.\n\nCapsular serotyping was done using the Quellung reaction test according to (Habib et al., 2014). Frozen vials containing Streptococcus pneumoniae stocks stored at -70°C were thawed at room temperature for about 30 minutes. Next, 10µl of the stored Streptococcus pneumoniae cells were suspended in 50µl PBS and gently vortexed. Subsequently, 10µl of the suspended cells were added on to a glass slide and mixed with 10µl pooled antisera (Statens Serum Institute, cat. No.16744). The glass slide was swirled gently while observing for any agglutination reaction until a positive reaction was observed with various pooled antisera. The process was repeated with individual groups under various antisera pools. After that, 10µl of the suspended cells in PBS were added to a glass slide and mixed with various Streptococcus pneumoniae serotype-specific antisera included in the antisera pools that gave a positive reaction. This was done until a positive reaction with the particular serotype specific antisera was observed.\n\nThose serotypes that did not belong to any pool were typed directly until a positive agglutination reaction was observed. The cells/PBS/serotype-specific antisera mixture on the glass slide were covered with a cover slip and observed under a phase contrast microscope with a ×400 objective lens with oil emulsion.\n\nAntimicrobial susceptibility testing (AST) was performed using the Kirby–Bauer disc diffusion method according to Clinical and Laboratory Standards Institute (CLSI) guideline (Hegstad et al., 2014). A 0.5 McFarland standard of freshly subcultured pneumococci organisms was inoculated on a 150 mm Mueller–Hinton plate with 5 % sheep blood (MH-BA). A standard disk dispenser was used to dispense the various antibiotic disks on the MH-BA agar (CLSI, 2004). Incubation was done overnight (20–24 hours) at 35°C in 5% CO2 and various clearance zones read with the aid of a vernier caliber. Antimicrobial agents tested included vancomycin (30µg, ≥17mm), erythromycin (15µg, ≥21mm), clindamycin (2µg, ≥19mm), oxacillin (1µg, ≥19mm) and ceftriaxone (1µg, ≥30mm). Because of ambiguity caused by poorly resolved concentration gradient around disks containing cephalosporins, this method may not give accurate results for ceftriaxone.\n\nData were entered, cleaned and analyzed on SPSS version 22. Descriptive statistics were used to summarize socio-demographic and antibiotic susceptibility patterns of Streptococcus pneumoniae isolates.\n\nAdjusted logistic regression was used to establish the strength of relationship between selected risk factors and susceptibility patterns of the selected antibiotic agents studied.\n\n\nResults\n\nA total of 206 subjects were recruited to participate in this study. It was found that 98% (n=202) of the subjects’ mothers/guardians were non-smokers as compared to 2% (n=4) who were smokers. About 96% (n=197) of the subjects did not share a household with an alcoholic, while 4% (n=9) did. Regarding housing, 55% (n=113) of the subjects lived in single rooms, 20% (n=42) lived in one-bedroom houses, 22% (n=46) lived in two-bedroom houses, while 2% (n=5) lived three or more bed roomed houses. It was found that 46% (n=94), 30% (n=62), 18% (n=38), 4% (n=9) and 1.5% (n=3) of the subjects cooked using gas, stove, charcoal, firewood and electricity, respectively. In addition, 75% (n=154) of the subjects’ households disposed of their waste in the public sewer system, 23% (n=47) disposed of their waste in private septic tanks and 2% (n=5) disposed of their waste using other undisclosed ways. The majority, 90% (n=185) of the subjects lived in households that used electricity as their source of light, while 5% (n=11) used traditional lamps, 2% (n=5) used candles for lighting, 1% (n=3) used solar energy and 0.49% (n=1) used lanterns. Within the two weeks preceding their visit to the clinic, 54% (n=112) of the subjects had consumed antibiotics, while 46% (n=94) had not. Daycare centers were not attended by 81% (n=167) of the subjects as compared to 19% (n=39) who did attend. Approximately 30% (n=61) of the subjects shared their households with three people other than their parents/guardians, 25% (n=52) shared with two other people, 18% (n=37) shared with one other person, 16% (n=33) shared with four other people, 6% (n=12) shared with five other people and 5% (n=10) shared with more than five other people. Lastly, 65% (n=133) of the subjects were exclusively breastfed, 35% (n=72) were moderately breastfed, while 0.49% (n=1) had never been breastfed (Table 1).\n\nn, total number of subjects per category from a total of 206; %, percentage of subjects per category.\n\nFrom the 206 subjects recruited, a total of 42 pneumococci were isolated. Out of the 42 total Streptococcus pneumoniae isolates, 39 (92.86%) were susceptible to erythromycin and vancomycin, eight (19.86%) were susceptible to oxacillin, 40 (95.24%) were susceptible to clindamycin and 24 (57.86%) were susceptible to ceftriaxone (Table 2). All the eight (19.05%) isolates of serotype 28F were susceptible to erythromycin, vancomycin and clindamycin; one isolate (2.38%) was susceptible to oxacillin and three (7.14%) were susceptible to ceftriaxone. All the five (11.9%) isolates of serotype 6A were susceptible to erythromycin; four (9.52%) were susceptible to vancomycin, one (2.38%) was susceptible to oxacillin, four (9.52%) were susceptible to clindamycin and two (4.76%) were susceptible to ceftriaxone.\n\nThe table above summarises the susceptibility of Streptococcus pneumoniae to various antibiotic agents at different concentrations. Vancomycin (30µg, ≥17mm); erythromycin (15µg, ≥21mm); clindamycin (2µg, ≥19mm); oxacillin (1µg, ≥19mm) and cetriaxone (1µg, ≥30mm) (CLSI, 2017).\n\nIsolates from serotypes 13, 15B, 21, 35B, 35F, 39, 48 and untypeable were susceptible to erythromycin, vancomycin, clindamycin and ceftriaxone but were non-susceptible to oxacillin, except serotype 21 and untypeable. Isolates from serotypes 3, 23A, 23B, 20, 11A, 17F, 19B and 7C were susceptible to erythromycin, vancomycin and clindamycin. Serotypes 3, 23A and 11A were resistant to erythromycin and serotypes 11A and 7C were resistant to vancomycin. Serotypes 3, 23A, 23B, 20, 17F, 19B and 7C were non-susceptible to oxacillin. Serotype 17F was resistant to clindamycin, while serotypes 3, 23B, 11A, 17F and 7C were resistant to ceftriaxone (Table 3).\n\nThe table above shows Streptococcus pneumoniae serotypes susceptibility profiles to selected antimicrobial agents. The antimicrobial agents are vancomycin (30µg, ≥17mm), erythromycin (15µg, ≥21mm), clindamycin (2µg, ≥19mm), oxacillin (1µg, ≥19mm) and cetriaxone (1µg, ≥30mm).\n\nCompared with males, Streptococcus pneumoniae isolated from female subjects were more likely to be resistant to erythromycin (OR 1.94; CI 0.23, 16.10) but odds of being resistant to vancomycin, oxacillin, clindamycin and ceftriaxone decreased when the subject was female: OR 0.14 (95% CI 0.01, 2.81), OR 0.51 (95% CI 0.11, 2.26), OR 0.51 (95% CI 0.11, 2.26), OR 0.2 (95% CI 0.01, 4.43) and OR 0.56 (95% CI 0.17, 1.86), respectively. The odds of being resistant to erythromycin, vancomycin, oxacillin, clindamycin and ceftriaxone decreased when the subject was aged 13–24 months, 25–36 months and 37–48 months old, but the odds increased when the subject was aged 49–60 months old. The odds of being resistant to erythromycin, vancomycin, oxacillin and clindamycin increased when the subjects had consumed antibiotics two weeks prior to the study. However, the odds of being resistant to ceftriaxone decreased. The odds of being resistant to the tested range of antimicrobial agents decreased when the subjects had had none, moderate and exclusive breastfeeding types with the exception of oxacillin whose odds of resistance increased when the subjects had had moderate breastfeeding. Odds of being resistant to the tested antimicrobial agents decreased when the subjects had received a full dose of PCV-10, except for vancomycin and ceftriaxone whose odds of resistance increased when the subjects had received a full dose of PCV-10. The odds of being resistant to the tested antimicrobial agents significantly increased when the subjects had been attending daycare centers (Table 4).\n\nOdds ratio (OR)=1: exposure does not affect odds of outcome; OR>1: exposure associated with higher odds of outcome; OR<1: exposure associated with lower rates of outcome. Confidence intervals (CI) are used to estimate the precision of the OR. The higher the CI the lower the precision of the OR and the lower the CI the higher the precision of the OR. P-value is a measure of the significance level of the exposure.\n\n\nDiscussion\n\nAntimicrobial resistance is a major global public health hazard in the waiting lounge. Infections that are resistant to antibiotics are often hard to treat, the cost of managing them is much higher and they generally pose more stress on families’ finances. Its envisaged that by 2050, general global mortality due to antimicrobial resistance will be at 10 million people annually (de Kraker et al., 2016); Streptococcus pneumoniae is likely to be one of the major culprits, the need for its earlier containment has never been more urgent.\n\nThe WHO (2014) recommends the use of penicillins as the first line of antibiotic therapy for treatment of pneumococcal disease and; clindamycin, erythromycin, vancomycin and cephalosporins as alternative antibiotic interventions. In the current study, most (93%, n=39) of the Streptococcus pneumoniae isolates were sensitive to erythromycin and vancomycin. Erythromycin and vancomycin act by inhibiting protein synthesis and development of the cell wall in the target bacteria (Tenson et al., 2003). Streptococcus pneumoniae tolerance to erythromycin and vancomycin can be attributed to the previous irregular exposure to the drug agents (Tenson et al., 1997).\n\nHowever, in the current study, the number of erythromycin and vancomycin tolerant Streptococcus pneumoniae is almost medically insignificant as compared to previous studies, which reported that rates of nasopharyngeal Streptococcus pneumoniae susceptibility to erythromycin and vancomycin were even higher among healthy preschool children in Vilnius (90.4 and 98 % in 2006, respectively) (Heelan et al., 2004).\n\nAbout 19% (n=8) of the Streptococcus pneumoniae isolates were susceptible to oxacillin. Being an antibiotic in the penicillin group of drugs, it is expected that it will be one of the most prescribed and irregularly used on the local market. The current results are similar to those observed among non-invasive Streptococcus pneumoniae strains in pediatric populations in Poland and Russia where only 29.5% to 29.2% and 16.7% to 19.3% of the isolates in these respective countries were susceptible to penicillins over a period of two consecutive years (Albrich et al., 2004; Mayanskiy et al., 2014) before routine PCV vaccination. What is common in both scenarios is that with poorly enforced National Antibiotic Use Plans (NAUP), the general population is predisposed to purchase of antibiotics over the counter without valid a physician’s prescription, effectively exposing the general populace to the dangers of resistant pneumococcal infections. This, coupled up with failure to consume the correct regimen, results in antibiotic selection pressure, which consequently brings forth resistant strains that are then horizontally spread across the population (Neu, 1992). This would explain the unexpected high resistance levels of Streptococcus pneumoniae serotypes to oxacillin. These results are consistent with those of Bronzwaer et al. (2002), who reported increased (53%, n=115) antibiotic failure, especially to different derivatives of penicillins.\n\nHigh levels of sensitivity and tolerance to clindamycin and ceftriaxone agree with findings by scientists at the St. Jude Children’s Research Hospital, who demonstrated that clindamycin and azithromycin, which kill bacteria by inhibiting protein synthesis, are more effective than a standard first-line treatment with the beta-lactam antibiotic ampicillin, which causes the bacteria to lyse, or burst (Sutcliffe et al., 1996).\n\nClindamycin is especially recommended for Streptococcus pneumoniae treatment because of its spectrum of activity includes group A streptococci (Puopolo et al., 2007). Usually, it is reserved for cases of Streptococcus pneumoniae with high resistance to penicillins and/or cases of subjects being allergic to first line antibiotic therapies (Karlström et al., 2009). The high level of ceftriaxone tolerant Streptococcus pneumoniae (CT- Streptococcus pneumoniae) observed in this study is concerning. This is because broad spectrum cephalosporins like ceftriaxone are recommended for invasive pneumococcal disease in cases where other first line antibiotics have failed (Molyneux et al., 2011). Resistance would only be expected in cases where the subjects had been inappropriately exposed to other cephalosporins as the first line of therapy (Von Mollendorf et al., 2014). Resistance to penicillins has also been reported to have a nexus with resistance to cephalosporins as they both target the beta-lactam ring of the bacteria (Drawz & Bonomo, 2010). Irregular and non-prescribed purchase of antibiotics at local pharmacies and shops may be the other culprits behind these unfortunate findings.\n\nPneumococcal resistance to ceftriaxone was found to be high (up to 66–67%) in Estonia (Tadesse et al., 2017) and Romania, while the rates were lower in Norway (4.6%) (Steens et al., 2015), the Netherlands (12.9%) (Voets et al., 2012) and the Czech Republic (15.7 %) (Zemlickova et al., 2010). Streptococcus pneumoniae has over 90 serotypes belonging to different groups according to their capsular polysaccharides and the degree of cross-reactivity (Hathaway et al., 2012). The various Streptococcus pneumoniae serotypes behave differently when exposed to different antibiotic agents (Reinert, 2009). In this study, seven serotype 28F strains out of eight were tolerant to oxacillin, while only three were sensitive to ceftriaxone. Oxacillin, which is a penicillin, is normally given as a first line of treatment for the various forms of pneumococcal disease infections and ceftriaxone is a broad spectrum antibiotic normally given after penicillin has failed (File, 2006). The mode of resistance to the two antimicrobial agents is shared to some extent as both involve the use of a four atom beta-lactam ring (Kong et al., 2010). As such, resistance in one is likely to influence resistance in another. Inappropriate exposure to one or both of these agents prior to the collection of the study samples may explain the high resistance levels in both.\n\nMost strains of serotypes 6A, 3, 23A, 23B, 20, 11A and 17 were oxacillin and ceftriaxone tolerant but highly susceptible to erythromycin, vancomycin and clindamycin. These results agree with findings from a previous study by Liñares et al. (2010), who demonstrated Streptococcus pneumoniae serotype susceptibility patterns to different antibiotics and intimated that serotype 28F and 6A are susceptible to erythromycin, vancomycin and clindamycin but highly resistant to penicillin and ceftriaxone. Serotypes 19A, 7C, 13, 15B, 21, 35B, 35F, 39, 48, and the untypeable Streptococcus pneumoniae were highly intolerant to erythromycin, clindamycin, vancomycin, oxacillin and ceftriaxone. While most Streptococcus pneumoniae serotypes in this study were susceptible to majority of the antimicrobials studied, resistance to oxacillin and ceftriaxone was clinically significant and largely worrying. A similar study examined antibiotic susceptibility of non-invasive Streptococcus pneumoniae in children aged less than six years with signs of an acute respiratory tract infection in Moscow, Russia. Non-susceptibility to penicillin, erythromycin and MDR was found in 28%, 26% and 22% of pneumococcal strains, respectively (Mayanskiy et al., 2014).\n\nThe slightly lower rates of Streptococcus pneumoniae resistance may be due to the enrolment of both vaccinated unvaccinated children and differences in the enactment and implementation of the antibiotic use/consumption policies. Although none of the risk factors had a significant influence on the Streptococcus pneumoniae nasopharyngeal carriage, odds demonstrated that age, gender, recent consumption of antibiotics by subjects, attendance at daycare centers, breast-feeding type and vaccination status had a bearing on the effectiveness of antibiotics. For instance, being female and young increased the subject’s susceptibility to erythromycin tolerance, while decreasing susceptibility to vancomycin, clindamycin, ceftriaxone and oxacillin. Children exposed to antibiotics within the two weeks prior to sample collection and PCV-10 unvaccinated children had increased odds of being resistant to all the antibiotics studied. Breastfeeding type, attendance at day-care centers and PCV-10 vaccination status play a major role in antibiotic susceptibility according to the findings of this study.\n\nWith the exception of oxacillin, the odds of Streptococcus pneumoniae being resistant to clindamycin, vancomycin, erythromycin, ceftriaxone decreased when subjects had been moderately or exclusively breastfed. According to Lynch & Zhanel (2009), due to the likelihood of horizontal transfer of Streptococcus pneumoniae resistant genes within a community, subjects who attended daycare centers and had not received PCV-10 immunization had increased chances of being resistant to all the sampled antibiotics. These results are consistent with findings from Lynch & Zhanel (2010) who postulated that age, PCV7 vaccination, attendance at day care or school, previous respiratory infection and non-susceptibility to penicillin were significantly or insignificantly associated with antibiotic resistance. While some of the results may not have strong research significance, they pose significant clinical concerns.\n\n\nData availability\n\nThis project contains the following underlying data:\n\nName of Repository: Harvard dataverse\n\nData Name: Streptococcus Pneumoniae Serotype Epidemiology, Antibiotic Susceptibility Profiles and Associated Risk Factors among Children Attending Getrudes Children’s Hospital\n\nDOI: https://doi.org/10.7910/DVN/VDMKS2 (Walekhwa, 2019)\n\nThis project contains the following extended data:\n\nQuestionnaire", "appendix": "Acknowledgements\n\nSpecial thanks to Ms. Angela Karani of Kemri Wellcome Trust, Kilifi for her professional input. The laboratory and technical team: Ann Karanu, Elizabeth Mbithe, Shadrack Mutua and Alfred Too. Your technical input was amazing! The Gertrude's Hospital clinical and laboratory team (Linet Okose, Barasa Kasili, Charles Muia, Lilive Njagi, Carolyne Thumbi, and Peninah Chege) you made this possible. Thanks to Japheth Wambani Rapando for the review of my work.\n\n\nReferences\n\nAlbrich WC, Monnet DL, Harbarth S: Antibiotic selection pressure and resistance in Streptococcus pneumoniae and Streptococcus pyogenes. Emerg Infect Dis. 2004; 10(3): 514–517. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAssamala: African Regional Health Report 2014. World Health Organisation. 2014; 187. Reference Source\n\nBronzwaer SL, Cars O, Buchholz U, et al.: A European study on the relationship between antimicrobial use and antimicrobial resistance. Emerg Infect Dis. 2002; 8(3): 278–282. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCLSI: Performance Standards for Antimicrobial Disk and Dilution Susceptibility Tests for Bacteria Isolated from Animals; Approved Standard; Informational Supplement. Clinical and Laboratory Standards Institute. 2004. Reference Source\n\nCLSI: Performance standards for antimicrobial susceptibility testing. 27th ed. CLSI supplement M100. Wayne, PA: Clinical and Laboratory Standards Institute. Performance standards for antimicrobial susceptibility testing. 27th ed. CLSI supplement M100. Wayne, PA: Clinical and Laboratory Standards Institute. 2017. Reference Source\n\nCole J: Antimicrobial resistance--a 'rising tide' of national (and international) risk. J Hosp Infect. 2016; 92(1): 3–4. PubMed Abstract \| Publisher Full Text\n\nde Kraker ME, Stewardson AJ, Harbarth S: Will 10 Million People Die a Year due to Antimicrobial Resistance by 2050? PLoS Med. 2016; 13(11): e1002184. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDonkor ES, Badoe EV: Insights into Pneumococcal Pathogenesis and Antibiotic Resistance. Adv Microbiol. 2014; 04(10): 627–643. Publisher Full Text\n\nDrawz SM, Bonomo RA: Three decades of beta-lactamase inhibitors. Clin Microbiol Rev. 2010; 23(1): 160–201. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFeldman C, Anderson R: Epidemiology, virulence factors and management of the pneumococcus [version 1; peer review: 2 approved]. F1000Res. 2016; 5: 2320. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFile TM Jr: Clinical implications and treatment of multiresistant Streptococcus pneumoniae pneumonia. Clin Microbiol Infect. 2006; 12 Suppl 3: 31–41. PubMed Abstract \| Publisher Full Text\n\nHabib M, Porter BD, Satzke C: Capsular serotyping of Streptococcus pneumoniae using the Quellung reaction. J Vis Exp. 2014; (84): e51208. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHathaway LJ, Brugger SD, Morand B, et al.: Capsule type of Streptococcus pneumoniae determines growth phenotype. PLoS Pathog. 2012; 8(3): e1002574. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHeelan JS, Hasenbein ME, McAdam AJ: Resistance of group B streptococcus to selected antibiotics, including erythromycin and clindamycin. J Clin Microbiol. 2004; 42(3): 1263–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHegstad K, Giske CG, Haldorsen B, et al.: Performance of the EUCAST disk diffusion method, the CLSI agar screen method, and the Vitek 2 automated antimicrobial susceptibility testing system for detection of clinical isolates of Enterococci with low- and medium-level VanB-type vancomycin resistance: a multicenter study. J Clin Microbiol. 2014; 52(5): 1582–1589. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHenriques-Normark B, Tuomanen EI: The pneumococcus: epidemiology, microbiology, and pathogenesis. Cold Spring Harb Perspect Med. 2013; 3(7): pii: a010215. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKarlström A, Boyd KL, English BK, et al.: Treatment with protein synthesis inhibitors improves outcomes of secondary bacterial pneumonia after influenza. J Infect Dis. 2009; 199(3): 311–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKong KF, Schneper L, Mathee K: Beta-lactam antibiotics: from antibiosis to resistance and bacteriology. APMIS. 2010; 118(1): 1–36. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLaxminarayan R, Duse A, Wattal C, et al.: Antibiotic resistance-the need for global solutions. Lancet Infect Dis. 2013; 13(12): 1057–98. PubMed Abstract \| Publisher Full Text\n\nLiñares J, Ardanuy C, Pallares R, et al.: Changes in antimicrobial resistance, serotypes and genotypes in Streptococcus pneumoniae over a 30-year period. Clin Microbiol Infect. 2010; 16(5): 402–10. PubMed Abstract \| Publisher Full Text\n\nLowth M: Pneumococcal disease. Practice Nurse. 2015. Reference Source\n\nLynch JP 3rd, Zhanel GG: Streptococcus pneumoniae: does antimicrobial resistance matter? Semin Respir Crit Care Med. 2009; 30(2): 210–38. PubMed Abstract \| Publisher Full Text\n\nLynch JP 3rd, Zhanel GG: Streptococcus pneumoniae: epidemiology and risk factors, evolution of antimicrobial resistance, and impact of vaccines. Curr Opin Pulm Med. 2010; 16(3): 217–25. PubMed Abstract\n\nMayanskiy N, Alyabieva N, Ponomarenko O, et al.: Serotypes and antibiotic resistance of non-invasive Streptococcus pneumoniae circulating in pediatric hospitals in Moscow, Russia. Int J Infect Dis. 2014; 20: 58–62. PubMed Abstract \| Publisher Full Text\n\nMitchell AM, Mitchell TJ: Streptococcus pneumoniae: virulence factors and variation. Clin Microbiol Infect. 2010; 16(5): 411–8. PubMed Abstract \| Publisher Full Text\n\nVon Mollendorf C, Cohen C, de Gouveia L, et al.: Factors associated with ceftriaxone nonsusceptibility of streptococcus pneumoniae: analysis of south african national surveillance data, 2003 to 2010. Antimicrob Agents Chemother. 2014; 58(6): 3293–305. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMolyneux E, Nizami SQ, Saha S, et al.: 5 versus 10 days of treatment with ceftriaxone for bacterial meningitis in children: a double-blind randomised equivalence study. Lancet. 2011; 377(9780): 1837–45. PubMed Abstract \| Publisher Full Text\n\nNeu HC: The crisis in antibiotic resistance. Science. 1992; 257(5073): 1064–73. PubMed Abstract \| Publisher Full Text\n\nO’Neill J: Tackling drug-resistant infections globally: final report and recommendations. The Review on Antimicrobial Resistance. 2016; 84. Reference Source\n\nPuopolo KM, Klinzing DC, Lin MP, et al.: A composite transposon associated with erythromycin and clindamycin resistance in group B Streptococcus. J Med Microbiol. 2007; 56(Pt 7): 947–55. PubMed Abstract \| Publisher Full Text\n\nReardon S: Antibiotic resistance sweeping developing world. Nature. 2014; 509(7499): 141–2. PubMed Abstract \| Publisher Full Text\n\nReinert RR: The antimicrobial resistance profile of Streptococcus pneumoniae. Clin Microbiol Infect. 2009; 15 Suppl 3: 7–11. PubMed Abstract \| Publisher Full Text\n\nSteens A, Caugant DA, Aaberge IS, et al.: Decreased Carriage and Genetic Shifts in the Streptococcus pneumoniae Population After Changing the Seven-valent to the Thirteen-valent Pneumococcal Vaccine in Norway. Pediatr Infect Dis J. 2015; 34(8): 875–83. PubMed Abstract \| Publisher Full Text\n\nSutcliffe J, Tait-Kamradt A, Wondrack L: Streptococcus pneumoniae and Streptococcus pyogenes resistant to macrolides but sensitive to clindamycin: a common resistance pattern mediated by an efflux system. Antimicrob Agents Chemother. 1996; 40(8): 1817–24. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTadesse BT, Ashley EA, Ongarello S, et al.: Antimicrobial resistance in Africa: a systematic review. BMC Infect Dis. 2017; 17(1): 616. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTenson T, Lovmar M, Ehrenberg M: The mechanism of action of macrolides, lincosamides and streptogramin B reveals the nascent peptide exit path in the ribosome. J Mol Biol. 2003; 330(5): 1005–14. PubMed Abstract \| Publisher Full Text\n\nTenson T, Xiong L, Kloss P, et al.: Erythromycin resistance peptides selected from random peptide libraries. J Biol Chem. 1997; 272(28): 17425–30. PubMed Abstract \| Publisher Full Text\n\nTuti T, Agweyu A, Mwaniki P, et al.: An exploration of mortality risk factors in non-severe pneumonia in children using clinical data from Kenya. BMC Med. 2017; 15(1): 201. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVentola CL: The antibiotic resistance crisis: part 1: causes and threats. P T. 2015; 40(4): 277–283. PubMed Abstract \| Free Full Text\n\nVoets GM, Platteel TN, Fluit AC, et al.: Population distribution of Beta-lactamase conferring resistance to third-generation cephalosporins in human clinical Enterobacteriaceae in the Netherlands. PLoS One. 2012; 7(12): e52102. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWalekhwa M: Streptococcus pneumoniae SEROTYPE EPIDEMIOLOGY, ANTIBIOTIC SUSCEPTIBILITY PROFILES AND ASSOCIATED RISK FACTORS AMONG CHILDREN ATTENDING GERTRUDES CHILDREN’S HOSPITAL. Harvard Dataverse, V3, 2019. http://www.doi.org/10.7910/DVN/VDMKS2\n\nWHO: Antimicrobial resistance. Global Report on Surveillance. Bull World Health Organ. 2014; 383–394. Reference Source\n\nZemlickova H, Jakubu V, Urbaskova P, et al.: Serotype-specific invasive disease potential of Streptococcus pneumoniae in Czech children. J Med Microbiol. 2010; 59(Pt 9): 1079–83. PubMed Abstract \| Publisher Full Text" }	[ { "id": "55502", "date": "05 Nov 2019", "name": "Bartholomew N. Ondigo", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAbstract\nProvide N, number of children sampled.\n\nClearance zones read in what units of measurements?\n\nResults – This need to be tailored as per the objective of the manuscript: patterns of Streptococcus pneumoniae among PCV-10 vaccinated and unvaccinated children ≤5 years old at Gertrude's Children’s Hospital. However, the results are written focusing on day care centres.\n\nKey words – italicise scientific names.\nIntroduction\n\nParagraph 1- Repetition of globally and worldwide in the same sentence. The focus of this paragraph need to be antimicrobial resistance.\n\nParagraph 2 - Rewrite to introduce Streptococcus pneumonia as the causative agent of pneumococcal disease.\n\nParagraph 3 – Should focus briefly on the mechanisms of antimicrobial resistance.\n\nParagraph 4 – The problem and uniqueness of the study.\nMethods\nThis was a descriptive cross-sectional study conducted among children attending Gertrude’s Children’s Hospital (GCH) in Nairobi.\n\nEthical statement: This was obtained and approved from ….. “not given…”. This statement need to be rewritten to reflect: Written consent was sought from each child guardian/parent with each consent form indicating clearly the study purpose, any anticipated consequences of the research, the anticipated uses of the data, possible benefits of the study and possible harm, confidentiality of the data and the option to withdraw their children participation at any given time.\n\nStudy location: Provide reference for Joint Commission International (JCA).\n\nSPSS version 22 – source, city, state?\nResults\nConfirm if the 4 kids were smokers, and if smokers what was the criteria used.\n\nConsume antibiotics does not sound medically – better word to use is take antibiotics.\n\nHave headings for the several parts of the results focusing on:\n\nSocio-Demographic characteristics – a brief summary (Table 1).\n\nAntimicrobial susceptibility patterns of Streptococcus pneumoniae among PCV-10 vaccinated children ≤5 years (N =?) by selected antibiotic agents. Table 3 need to be split to show vaccinated children, then describe the results in text.\n\nAntimicrobial susceptibility patterns of Streptococcus pneumoniae among PCV-10 unvaccinated children ≤5 years (N=?) by selected antibiotic agents. Table 3 need to be split to show unvaccinated children, then describe the results in text.\n\nRisk factors and susceptibility patterns of vaccinated children by selected antibiotic agents.\n\nRisk factors and susceptibility patterns of unvaccinated children by selected antibiotic agents.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "66815", "date": "10 Aug 2020", "name": "Izabela Korona-Glowniak", "expertise": [ "Reviewer Expertise microbiology", "antibiotic resistance", "microbiological diagnostics" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript deals with an interesting topic in the field of pneumococcal resistance epidemiology. The link between antibiotic use and the emergence of bacterial resistance has been clearly established and many studies have shown the increase in antibiotic resistance of S. pneumoniae. Due to geographical diversity of S. pneumoniae resistance, which depends on the local antimicrobial policy, epidemiological studies in each geographical region should be determined and analyzed separately. The topic is interesting and worthy of note, yet the manuscript have to be improved.\n\nAbstract: Methods paragraph is too detailed.\n\nMethod: In my opinion, the methods - especially related to diagnostic tests - are described too extensively.\n\nResults: The authors should avoid duplicating the results in tables and in the text. The first paragraph of the Results is unnecessary. All data are shown in Table.\n\nThe last sentence of the Result \" The odds of being resistant to the tested antimicrobial agents significantly increased when the subjects had been attending daycare centers\" is an over interpretation of the statistical analysis. All but one tests showed insignificant differences between the analyzed groups.\n\nThe term odds in description of the statistical analysis results sounds awkward. The authors should rewrite the text using other terms, for example risk.\n\nDiscussion: The first paragraph of the Discussion is a repetition of information given in the Introduction.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-1699
https://f1000research.com/articles/8-1695/v1	27 Sep 19	{ "type": "Research Article", "title": "Discovery of novel plasma biomarker ratios to discriminate traumatic brain injury", "authors": [ "Michelle Chen", "Antoninus Soosaipillai", "Douglas D. Fraser", "Eleftherios P. Diamandis", "Michelle Chen", "Antoninus Soosaipillai", "Douglas D. Fraser" ], "abstract": "Background: Traumatic brain injury (TBI) is a major cause of death and disability. Despite increased awareness, reliable biomarkers are urgently needed to aid in all forms of traumatic brain injury diagnosis and prognosis. Methods: Here, we aim to assess the diagnostic utility of known and novel TBI biomarkers in a pilot patient cohort of severe TBI (sTBI) patients and healthy controls. We analyzed concentrations of S100 calcium binding protein B (S100B), neuron specific enolase (NSE), human kallikrein 6 (hK6) and prostaglandin D2 synthase (PGDS) using ELISA immunoassays. Results: Plasma levels of hK6 and PGDS were significantly lower in sTBI compared with controls, while S100B and NSE were significantly higher. Furthermore, we show that ratios of NSE and S100B with hK6 and PGDS may be able to determine the presence of sTBI better than single markers alone. Conclusions: The findings presented here represent a starting point for future validation, where biomarker ratios can be tested in independent TBI cohorts.", "keywords": [ "Kallikrein 6", "prostaglandin D synthase", "concussion", "biomarker", "biomarker ratios", "traumatic brain injury" ], "content": "Abbreviations\n\nTBI: Traumatic brain injury, mTBI: mild traumatic brain injury, sTBI: severe traumatic brain injury, S100B: S100 calcium binding protein B, NSE: neuron specific enolase, hK6: human kallikrein 6, PGDS: prostaglandin D2 synthase, GCS: Glasgow Coma Scale, CT: computed tomography, UCH-L1: ubiquitin C-terminal hydrolase-L1, GFAP: glial fibrillary acidic protein, PPV: positive predictive value, ROC: receiver operating characteristic, AUC: area under the curve, CI: confidence interval, LOS: length of stay\n\n\nIntroduction\n\nTraumatic brain injury (TBI) is a significant cause of morbidity, with an annual incidence of over 65 million cases worldwide1. TBI is caused by trauma to the head or body that causes neurologic damage and dysfunction. It occurs most often due to falls, but is also a common occurrence in motor vehicle collisions, sports and assaults2. These mechanical injuries can vary in both the severity and the form of injury to the brain3,4. Generally, presenting TBI may be classified as mild, moderate, or severe. Patients are primarily stratified using the Glasgow Coma Scale (GCS), a scoring system that evaluates verbal performance, motor function and eye function. A lower GCS score indicates a more severe diagnosis and may therefore indicate the need for further investigations and possibly medical and/or surgical interventions.\n\nGiven the significant burden of disease, reliable diagnostic and prognostic approaches are urgently needed to guide care. One promising solution is the quantification of blood biomarkers capable of delineating the severity of trauma and/or prognostic outcome. These proteins enter the circulation due to disruption of brain structures such as the blood brain barrier and/or axons, leading to a neuroinflammatory response5,6. Although various markers of mTBI have been identified, a validated panel has yet to be resolved7. For example, S100 calcium-binding protein B (S100B) and neuron-specific enolase (NSE) are two well-documented biomarkers with good correlations to TBI outcome. However, their performance may be limited by non-specific correlations8,9. The Food and Drug Administration recently approved the first blood biomarker test to guide head imaging after TBI; it measures ubiquitin C-terminal hydrolase-L1 (UCH-L1) and glial fibrillary acidic protein (GFAP)10. Despite the approval, concerns have been raised regarding the relatively low positive predictive value (PPV) of the test and its utility as a marker for TBI diagnosis specifically11. Given this, it is clear that there is a need for improved diagnostic TBI biomarkers with validated efficacy in a clinical setting.\n\nHuman kallikrein 6 (hK6) is a serine protease with high brain and spinal cord expression, particularly in oligodendrocytes12,13. Previous studies have linked hK6 to brain-related disorders such as spinal cord injury, where it may play a role in myelin repair14,15. Furthermore, work by our group has demonstrated that serum hK6 may have significant clinical utility as a prognostic marker of non-traumatic subarachnoid hemorrhage16. Prostaglandin D synthase (PGDS) is a highly expressed enzyme that represents nearly 3% of the total protein in cerebral spinal fluid17. The role of PGDS in hypoxic-ischemic brain injury has been well-characterized18,19. Therefore, we hypothesize that hK6 and PGDS may be novel biomarkers of TBI and may have predictive value in its diagnosis.\n\nIn this study, we measured plasma concentrations of multiple proteins in sTBI patients compared with healthy controls. These markers include those well established in literature (S100B and NSE), as well as those we hypothesize to be detectable in plasma following TBI – hK6 and PGDS. We reason that the concentration of these proteins will fluctuate following brain injury, causing them to be more or less abundant in TBI patient plasma compared with controls. Furthermore, we hypothesized that the combination of these proteins may be useful biomarkers to delineate the degree of TBI in patient populations.\n\n\nMethods\n\nPlasma samples were obtained using strict standard operating procedures from the Translational Research Centre, London, Ontario, Canada (https://translationalresearchcentre.com/). Study protocols were approved by The Western University Health Science Research Ethics Board (REB # 16693) and by The Mount Sinai Hospital Research Ethics Board (REB #18-0069-E). Informed consent was obtained either from the patient or a substitute decision maker.\n\nBlood samples were drawn from adults with severe TBI (sTBI; n=10), as defined by GCS ≤ 8 with abnormal CT findings, within 24 h of hospital admission. Samples from age-/sex-matched healthy controls (n=10) were also obtained from the Translational Research Centre. Blood was collected into 3%, 0.109 M sodium citrate tubes, centrifuged for 15 min (1500 x g at 4 °C), and the plasma was frozen at −70 °C. Detailed collection and handling protocols have been described previously20,21. Plasma samples were transferred on dry ice to Mount Sinai Hospital, Toronto, Ontario, Canada and stored at -80 °C until analysis.\n\nThe concentrations of S100B and NSE were measured using electrochemoluminometric immunoassay (Cobas e411 Immunoanalyzer; Roche Diagnostics, Penzberg, Germany) according to the manufacturer’s protocol, with a detection limit of 0.02 ng/mL and 0.19 ng/mL, respectively. All samples were thawed at room temperature and analyzed in one run.\n\nFor all hK6 measurements, samples were thawed at room temperature, then diluted 10 fold with 6% BSA (Wisent Bioproducts, Quebec, Canada). hK6 was analyzed with an in-house spectrophotometric ELISA assay with a detection limit of 0.078 ng/mL and dynamic range of 0.078–5 ng/mL22. Briefly, samples were analyzed in duplicate using an hK6 monoclonal capture antibody and mouse monoclonal biotinylated detection antibody. White polystyrene 96-well plates were coated with 400 ng of capture antibody in 100 μl of coating buffer (50 mmol/L Tris-HCl) and incubated overnight. Plates were then washed in wash buffer (10 mmol/L Tris-HCl buffer, 150 mmol/L NaCl and 0.5 mL/L Tween 20). 50 μl of diluted sample and 50 μl of green assay buffer (60 g/L BSA, 0.1 g/L goat globulin, 0.02 g/L mouse globulin, 1 g/L bovine globulin, 5 ml/L Tween 20 and 37 g/L KCl) were added to each well and incubated with shaking for 2 h. Plates were washed, incubated with 50 ng (100 μl) of detection antibody for 1 h, washed, and incubated again with 5 ng (100 μl) of alkaline phosphatase-conjugated streptavidin for 15 min. 100 μl of diflunisal phosphate solution (0.1 M NaOH containing 10 mM diflunisal phosphate) diluted 1:20 in substrate buffer (0.1 M Tris, 0.15 M NaCl, 1 mM MgCl2 and 0.05% NaN3) was added to each well and incubated for 10 min. The reaction was stopped with 100 µl/well of developing solution (1 M Tris, 0.15 M NaOH, 2 mM TbCl3 and 3 mM EDTA). Excitation at 340 nm and emission at 615 nm was measured with a time-resolved fluorometer (EnVision, Perkin-Elmer). Samples were diluted 200 fold in 6% BSA for PGDS analysis via similar in-house ELISA, with a detection limit of 0.2 ng/mL and dynamic range of 0.2–50 ng/mL, as described in detail elsewhere23.\n\nRatios of NSE/hK6, S100B/hK6, NSE/PGDS, S100B/PGDS were calculated for data analysis. Statistical analyses were conducted using GraphPad Prism version 6.0e. For receiver operating characteristic (ROC) analysis, the area under the curve (AUC) estimates with 95% confidence intervals (CIs) were calculated using DeLong’s method24. CIs were calculated using 2000 bootstrap replicates each. P-values were calculated using Wilcoxon signed-rank tests. Spearman correlations were conducted between biomarker value and lowest-documented GCS score (within 24 h of hospital admission), as well as hospital length of stay (LOS). Mann-Whitney U test was used to compare biomarker distribution in patients vs. controls. Two-tailed student’s t-test was used to compare biomarker concentrations between patient and control groups. One sTBI subject with extreme high values for all assays was omitted from data analysis due to potential sample contamination or processing error.\n\n\nResults\n\nDemographics and pathological characteristics of recruited patients are summarized in Table 1. All raw clinical and biomarker data are available as Underlying data25.\n\nGCS, Glasgow Coma Score; ICU, intensive care unit; LOS, length of stay.\n\nMedian (SD).\n\n†Within 24 h of hospital admission.\n\nWe used in-house ELISAs and electrochemoluminometric assays to measure biomarker concentrations in the plasma of adult sTBI patients and matched healthy controls. Plasma levels of hK6 and PGDS were significantly lower (p = 0.0076 and p = 0.0172, respectively) in sTBI compared with controls, while S100B and NSE were significantly higher (p = 0.0002 and p = 0.0002, respectively) (Figure 1). All analyzed biomarker ratios (NSE/hK6, S100B/hK6, NSE/PGDS, S100B/PGDS) showed significant differences in sTBI vs control (p < 0.0001, for all; Figure 2).\n\nDotted lines represent group median, bars show interquartile range. Mann Whitney U test; p < 0.05, p < 0.01, p < 0.001, ns not significant.\n\nDotted lines represent group median, bars show interquartile range. Mann Whitney U test; ***p < 0.0001.\n\nIndividual biomarkers and calculated ratios were evaluated to determine their ability to discriminate between sTBI and control groups. AUC estimates and CIs are presented in Table 2. Plasma concentrations of NSE, S100B, hK6, PGDS and all ratios had significant predictive ability to discriminate sTBI from healthy controls (Figure 3, Table 2). Biomarker ratios showed the highest AUC values (AUC = 1.0, for all), while NSE and S100B (AUC = 0.967 for both) performed better than hK6 (AUC = 0.856) and PGDS (AUC = 0.822).\n\nThere were no significant correlations between any individual biomarker values and lowest-documented GCS score (Extended data: Supplemental Figures S1, Table S126). However, ratios of S100B/hK6 (rs = 0.7353, p = 0.0292) and S100B/PGDS (rs = 0.7899, p = 0.0151) were significantly associated with GCS score (Figure 4). Finally, we did not observe any significant differences in biomarker or ratio levels between sTBI survivors and non-survivors (Extended data: Supplemental Figures S2 and S326).\n\nSpearman’s correlation: (a) S100B/hK6; rs = 0.7353, p = 0.0292; (b) S100B/PGDS; rs = 0.7899, p = 0.0151; (c) NSE/hK6; ns; (d) NSE/PGDS; ns.\n\nPlasma concentrations of biomarkers and ratios were compared to hospital length of stay (LOS) of surviving sTBI patients. LOS was significantly correlated with hK6 (rs = -0.9276, p < 0.006) and PGDS (rs = -0.9276, p < 0.006), with lower biomarker levels corresponding to longer LOS (Figure 5, non-significant data not shown).\n\nSpearman’s correlation: (a) hK6; rs = -0.9276, p < 0.006; (b) PGDS; rs = -0.9276, p < 0.006.\n\n\nDiscussion\n\nIn this pilot study, we measured plasma concentrations of multiple proteins in sTBI patients compared with age-/sex-matched healthy controls. We analyzed concentrations of S100B, NSE, kallikrein 6 (hK6) and prostaglandin D2 synthase (PGDS). We show that ratios of NSE and S100B with hK6 and PGDS may be able to determine the presence of sTBI with 100% accuracy.\n\nPlasma levels of hK6 and PGDS were significantly lower in sTBI compared with controls, while S100B and NSE were significantly higher (Figure 1). This is consistent with abundant research demonstrating increased S100B and NSE post-TBI27–29. PGDS has also been shown to be reduced in the CSF of pediatric patients with inflicted TBI30. To our knowledge, our findings linking hK6 with sTBI have not been previously reported.\n\nInterestingly, the ratio of NSE/PGDS was able to segregate sTBI patients from healthy controls better than any marker alone. We observed one sTBI patient with NSE and S100B values that fell within the control range, but had a NSE/PGDS and S100B/hK6 value that was clearly higher than controls (Figure 1 and Figure 2). This suggests that these ratios may provide better sensitivity in determining the presence of sTBI. Furthermore, ROC analysis shows higher discriminatory capability in all ratios analyzed (AUC = 1.00, p = 0.0000217) in comparison with S100B and NSE alone (AUC = 0.967, p = 0.00015 for both, Figure 3). Taken together, our data suggest that the use of known TBI biomarkers, in conjunction with PGDS and hK6, may have significant diagnostic capability.\n\nAlthough we did not identify any significant correlations between individual biomarker concentrations and lowest documented GCS score (Extended data: Supplemental Figures S126), we found that ratios of S100B/hK6 and S100B/PGDS were significantly correlated with TBI severity. We observe that higher ratio values are associated with lower GCS score in sTBI patients (Figure 4). Furthermore, plasma concentrations of hK6 and PGDS were significantly and negatively associated with hospital LOS (Figure 5). Further investigations with a larger sample size are needed to confirm the prognostic utility of these novel findings.\n\nThe data presented here represents a pilot study exploring the utility of four biomarkers and their ratios in diagnosing sTBI. Overall, it appears that novel ratios of S100B, NSE, PGDS and kallikrein 6 may have improved TBI diagnostic capability over any one of these markers alone and may additionally have prognostic utility. Given the small sample size, these results must be validated in a larger cohort of TBI patients. Results from these experiments may provide valuable insight into TBI diagnosis and prognosis.\n\n\nData availability\n\nHarvard Dataverse: Underlying data: Discovery of novel plasma biomarker ratios to discriminate traumatic brain injury. https://doi.org/10.7910/DVN/VMR7MS25.\n\nThis project contains the following underlying data:\n\n- biomarker_data (biomarker data, including plasma concentrations of protein biomarkers and ratio values).\n\n- clinical_data (clinical data, including demographic information and clinical presentation data for each enrolled patient).\n\nHarvard Dataverse: Extended data: Discovery of novel plasma biomarker ratios to discriminate traumatic brain injury. https://doi.org/10.7910/DVN/UOOSXE26.\n\nThis project contains the following extended data:\n\n- Supplemental Figure S1. Individual biomarker correlation with lowest documented Glasgow Coma Scale (GCS) score in patients with sTBI (n=9).\n\n- Supplemental Figure S2. Differences in individual biomarker concentrations between sTBI patient outcomes (n=9).\n\n- Supplemental Figure S3. Differences in plasma biomarker ratios between sTBI patient outcomes (n=9).\n\n- Supplemental Table S1. Spearman’s correlation values between biomarkers and lowest documented GCS score in sTBI patients (n=9).\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "References\n\nDewan MC, Rattani A, Gupta S, et al.: Estimating the global incidence of traumatic brain injury. J Neurosurg. 2018; 1–18. PubMed Abstract \| Publisher Full Text\n\nHutchison MG, Lawrence DW, Cusimano MD, et al.: Head Trauma in Mixed Martial Arts. Am J Sports Med. 2014; 42(6): 1352–8. PubMed Abstract \| Publisher Full Text\n\nMckee AC, Daneshvar DH: The neuropathology of traumatic brain injury. Handb Clin Neurol. 2015; 127: 45–66. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSignoretti S, Lazzarino G, Tavazzi B, et al.: The pathophysiology of concussion. PM R. 2011; 3(10 Suppl 2): S359–68. PubMed Abstract \| Publisher Full Text\n\nShlosberg D, Benifla M, Kaufer D, et al.: Blood-brain barrier breakdown as a therapeutic target in traumatic brain injury. Nat Rev Neurol. 2010; 6(7): 393–403. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLoane DJ, Byrnes KR: Role of microglia in neurotrauma. Neurotherapeutics. 2010; 7(4): 366–77. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJones A, Jarvis P: Review of the potential use of blood neuro-biomarkers in the diagnosis of mild traumatic brain injury. Clin Exp Emerg Med. 2017; 4(3): 121–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPelinka LE, Hertz H, Mauritz W, et al.: Nonspecific increase of systemic neuron-specific enolase after trauma: clinical and experimental findings. Shock. 2005; 24(2): 119–23. PubMed Abstract \| Publisher Full Text\n\nAnderson RE, Hansson LO, Nilsson O, et al.: High serum S100B levels for trauma patients without head injuries. Neurosurgery. 2001; 48(6): 1255–1258; discussion 1258–60. PubMed Abstract \| Publisher Full Text\n\nBazarian JJ, Biberthaler P, Welch RD, et al.: Serum GFAP and UCH-L1 for prediction of absence of intracranial injuries on head CT (ALERT-TBI): a multicentre observational study. Lancet Neurol. 2018; 17(9): 782–9. PubMed Abstract \| Publisher Full Text\n\nLindberg DM: Finally, actual data for the FDA-approved biomarkers of traumatic brain injury. NEJM J Watch. 2018; [cited 2019 Jun 20]. Reference Source\n\nShaw JL, Diamandis EP: Distribution of 15 human kallikreins in tissues and biological fluids. Clin Chem. 2007; 53(8): 1423–32. PubMed Abstract \| Publisher Full Text\n\nTerayama R, Bando Y, Takahashi T, et al.: Differential expression of neuropsin and protease M/neurosin in oligodendrocytes after injury to the spinal cord. Glia. 2004; 48(2): 91–101. PubMed Abstract \| Publisher Full Text\n\nScarisbrick IA, Sabharwal P, Cruz H, et al.: Dynamic role of kallikrein 6 in traumatic spinal cord injury. Eur J Neurosci. 2006; 24(5): 1457–69. PubMed Abstract \| Publisher Full Text\n\nBando Y, Ito S, Nagai Y, et al.: Implications of protease M/neurosin in myelination during experimental demyelination and remyelination. Neurosci Lett. 2006; 405(3): 175–80. PubMed Abstract \| Publisher Full Text\n\nMartínez-Morillo E, Diamandis A, Romaschin AD, et al.: Kallikrein 6 as a serum prognostic marker in patients with aneurysmal subarachnoid hemorrhage. PLoS One. 2012; 7(9): e45676. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nClausen J: Proteins in normal cerebrospinal fluid not found in serum. Proc Soc Exp Biol Med. 1961; 107: 170–2. PubMed Abstract \| Publisher Full Text\n\nGonzalez-Rodriguez PJ, Li Y, Martinez F, et al.: Dexamethasone protects neonatal hypoxic-ischemic brain injury via L-PGDS-dependent PGD2-DP1-pERK signaling pathway. PLoS One. 2014; 9(12): e114470. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTaniguchi H, Mohri I, Okabe-Arahori H, et al.: Early induction of neuronal lipocalin-type prostaglandin D synthase after hypoxic-ischemic injury in developing brains. Neurosci Lett. 2007; 420(1): 39–44. PubMed Abstract \| Publisher Full Text\n\nGillio-Meina C, Cepinskas G, Cecchini EL, et al.: Translational research in pediatrics II: blood collection, processing, shipping, and storage. Pediatrics. 2013; 131(4): 754–66. PubMed Abstract \| Publisher Full Text\n\nDaley M, Dekaban G, Bartha R, et al.: Metabolomics profiling of concussion in adolescent male hockey players: a novel diagnostic method. Metabolomics. 2016; 12(12): 185. Publisher Full Text\n\nDiamandis EP, Yousef GM, Soosaipillai AR, et al.: Immunofluorometric assay of human kallikrein 6 (zyme/protease M/neurosin) and preliminary clinical applications. Clin Biochem. 2000; 33(5): 369–75. PubMed Abstract \| Publisher Full Text\n\nMelegos DN, Diamandis EP, Oda H, et al.: Immunofluorometric assay of prostaglandin D synthase in human tissue extracts and fluids. Clin Chem. 1996; 42(12): 1984–91. PubMed Abstract\n\nDeLong ER, DeLong DM, Clarke-Pearson DL: Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics. 1988; 44(3): 837–45. PubMed Abstract \| Publisher Full Text\n\nChen M, Soosaipillai A, Fraser DD, et al.: Underlying data: Discovery of novel plasma biomarker ratios to discriminate traumatic brain injury. Harvard Dataverse, V2, 2019. http://www.doi.org/10.7910/DVN/VMR7MS\n\nChen M, Soosaipillai A, Fraser DD, et al.: \"Extended data: Discovery of novel plasma biomarker ratios to discriminate traumatic brain injury\". Harvard Dataverse, V1, 2019. http://www.doi.org/10.7910/DVN/UOOSXE\n\nRodríguez-Rodríguez A, Egea-Guerrero JJ, Gordillo-Escobar E, et al.: S100B and Neuron-Specific Enolase as mortality predictors in patients with severe traumatic brain injury. Neurol Res. 2016; 38(2): 130–7. PubMed Abstract \| Publisher Full Text\n\nBerger RP, Pierce MC, Wisniewski SR, et al.: Neuron-specific enolase and S100B in cerebrospinal fluid after severe traumatic brain injury in infants and children. Pediatrics. 2002; 109(2): E31. PubMed Abstract \| Publisher Full Text\n\nBöhmer AE, Oses JP, Schmidt AP, et al.: Neuron-specific enolase, S100B, and glial fibrillary acidic protein levels as outcome predictors in patients with severe traumatic brain injury. Neurosurgery. 2011; 68(6): 1624–1630; discussion 1630–1. PubMed Abstract \| Publisher Full Text\n\nGao WM, Chadha MS, Berger RP, et al.: A gel-based proteomic comparison of human cerebrospinal fluid between inflicted and non-inflicted pediatric traumatic brain injury. J Neurotrauma. 2007; 24(1): 43–53. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "54418", "date": "18 Oct 2019", "name": "Joško Osredkar", "expertise": [ "Reviewer Expertise Laboratory medicine." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nTime for blood sampling after the brain injury is critical, especially for S100B. Biomarker S100B is increased within minutes after blood-brain-barrier (BBB) disruption. The passage to the systemic circulation is critical, especially the fact that only the BBB is damaged and not the brain. Another point is that most of these markers are expressed to varying degrees outside the central nervous system.\nIf possible a more precise time between injury and sampling would be an additional help in the interpretation. In the next study, send a sample within the specified time to calculate the difference in S100B concentrations.\nIn severe traumatic brain injury, we have cerebrospinal fluid (CSF) as well and it would be helpful to determine albumin (in CSF and plasma) to see if there is a brain or vessel damage.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "56390", "date": "12 Nov 2019", "name": "Firas H. Kobeissy", "expertise": [ "Reviewer Expertise Brain injury and biomarkers of TBI." ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nWe read with interest the study by Chen et al., where the authors have assessed (hK6) and (PGDS) as new markers of injury indicators coupled with NSE and S100B. These two proteins [(hK6) and (PGDS)] are of high interest and worth looking at; however, this study lacks enough power (n=10 exp vs 10 control) which is not enough to draw any kind of discriminatory level.\nN of patients should be increased and other demographic data should be included as well. The TBI patients assessed are in the severe level and as the authors have mentioned that both UCH-L1 and GFAP have been considered the gold standard by the FDA specifically for severe TBI which is the case in this study. This is in contrast to the S100B and NSE which have been criticized for their non-specificity (muscle and red blood cells derived!). Thus, I think these data should be contrasted to UCH-L1 and GFAP with higher N.\nIt is of interest as to why the CT scan was not included in the comparison.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "58587", "date": "23 Jan 2020", "name": "Marco Antonio Stefani", "expertise": [ "Reviewer Expertise Neurosurgery", "clinical and translational research." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting study carried out by Chen et al., focusing on human serum biomarkers values to improve the diagnosis and predict prognosis of severe traumatic brain injury (sTBI). They have described human kallikrein 6 (hK6) and prostaglandin D2 synthase (PGDS) as potential new markers of injury, coupled with NSE and S100B.\nThe methods are clear and the results are very well described. A larger sample is necessary to confirm the proposed ratios, as pointed out by the authors, so multiple variables can be combined, increasing the statistic power. In consequence, the conclusions are still limited, but these findings are promising and worthy of future research.\nThe pathological/biochemical mechanism to explain the increase of S100B and NSE in sTBI has been described and makes perfect sense. The protein S100B is produced and released primarily by astrocytes in the central nervous system and is a peripheral biochemical marker of neural injury, including reactive gliosis, astrocytic death, and/or blood-brain barrier dysfunction. Neuron-specific enolase (NSE) is a glycolytic pathway enzyme, predominantly neuronal, so increased CSF/serum levels of NSE and GFAP indicate neuronal and astrocyte damage.\nWe are still puzzled by the role of hK6 and PGDS and further research should focus also on describing the underlying cellular mechanisms related to the presented findings.\nIt is also noticeable that S100B and NSE proteins are also expressed in other non-neural cell types under physiological and pathological conditions, so future studies should control for that, especially in trauma.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1695
https://f1000research.com/articles/8-807/v2	25 Jun 19	{ "type": "Software Tool Article", "title": "InferAMP, a python web app for copy number inference from discrete gene-level amplification signals noted in clinical tumor profiling reports", "authors": [ "Paraic A. Kenny" ], "abstract": "As somatic next-generation sequencing gene panel analysis in advanced cancer patients is becoming more routine, oncologists are frequently presented with reports containing lists of genes with increased copy number. Distinguishing which of these amplified genes, if any, might be driving tumor growth and might thus be worth considering targeting can be challenging. One particular issue is the frequent absence of genomic contextual information in clinical reports, making it very challenging to determine which reported genes might be co-amplified and how large any such amplicons might be. We describe a straightforward Python web app, InferAMP, into which healthcare professionals may enter lists of amplified genes from clinical reports. The tool reports (1) the likely size of amplified genomic regions, (2) which reported genes are co-amplified and (3) which other cancer-relevant genes that were not evaluated in the assay may also be co-amplified in the specimen. The tool is accessible for web queries at http://inferamp.org.", "keywords": [ "cancer", "genetic testing", "copy number variation", "gene amplification", "oncology", "targeted therapy" ], "content": "Introduction\n\nFocal somatic gene copy number changes are a widespread event in tumor evolution1. Although these regions of amplification may be large, encompassing many hundreds of genes, typically only one or a small number of genes within the amplified regions are involved in driving tumor growth. Identification of the key driver genes within recurrent amplicons has led to the approval of some therapies that have changed clinical practice (e.g. anti-ERBB2 agents2); however, targeting other amplified genes such as FGFR family members3,4, EGFR5 or KIT6 has frequently proved disappointing. Nevertheless, even some of the more negative trials include occasional strong responses, indicating that sub-populations of patients with amplification of these oncogenes may experience clinical benefit if they can be identified.\n\nWith the goal of individualizing treatment for cancer patients, next-generation sequencing from tumor specimens is becoming widely adopted7. In addition to somatic point mutations, several of these assays report copy number changes in assayed genes. Reports for physicians typically present a list of amplified genes without providing a genomic context, leaving physicians and molecular tumor boards to hypothesize which of the listed genes might be driver genes suitable for therapeutic targeting. Given the poor response rates that have often been observed in clinical studies with amplified genes (compared to targeting genes activated by point mutation or fusion), physicians are often appropriately cautious about deciding whether a reported amplified gene may be actionable. Thus, many patients are spared receiving ineffective therapies, but a subgroup of patients who may experience clinical benefit do not get that opportunity.\n\nHere we provide an easy-to-use web tool for analyzing clinical genomics reports of amplified genes. It determines (1) the likely size of amplified genomic regions, (2) which reported genes are co-amplified and (3) which other cancer-relevant genes that were not evaluated in the assay may also be co-amplified in the specimen.\n\nThe primary goals are to allow healthcare professionals to determine whether the amplification region surrounding a particular oncogene is relatively small and lacking in other likely candidate cancer drivers (which may indicate increased likelihood that the analyzed gene is a driver) from larger amplicons with additional candidate driver genes (which would suggest a reduced probability that the reported gene is a driver). The approach was developed to analyze the widely used Foundation One test (329 genes) provided by Foundation Medicine but, by simply editing the target gene list, it can be generalized to tests from other vendors which report copy number variation throughout the genome.\n\n\nMethods\n\nInferAMP is written in Python 2.7 with Flask and implemented as a web service running on the Google App Engine (http://inferamp.org). Additional supplied requirements are (1) the coordinates of genes in the human genome, ‘coordinates.txt’ (hg38, UCSC genome browser), (2) The gene list from the assay of interest, ‘foundationone.txt’ and (3) a file listing genes recurrently altered in cancer from COSMIC8 (retrieved 5/3/2018), ‘cosmic.txt’.\n\nAn html page with a single query window allows the user to enter a comma-delimited list of genes reported as being amplified. The entry is passed to the script and parsed into individual gene names. An error check is performed to confirm that all entered gene names correspond to gene names in the genome used. The entered genes are considered to be amplified, while the other genes in the assay are considered to be not amplified.\n\nA simplified schema of how the algorithm works is presented in Figure 1, which depicts a chromosomal region containing 30 genes. Seven cancer-relevant genes are present, five of which are evaluated by the genomic assay (Figure 1A). In this test example, three genes were reported to be amplified (Figure 1B). Running the algorithm identifies these amplified genes (8, 13, 20) as well as the nearby assayed genes that are not reported amplified (4, 27). The algorithm considers all genes located between genes reported as amplified to also be amplified (Figure 1C, red shaded region). Because not every gene is assayed, precisely delineating the boundaries of an amplicon is not possible. To address this, the algorithm determines the nearest non-amplified gene at each end of the amplicon and infers that the genes located up to, but not including that gene may be possibly amplified (Figure 1D). The script then returns an html report page listing the entered genes, the amplicons into which they fall (in many cases, several discrete genes will be consolidated into a single amplicon), and also the other cancer-relevant genes within these regions that may be co-amplified with the reported genes. All genes reported include hyperlinks to that gene’s page on COSMIC.\n\n(A). Schematic diagram of a model genomic region with 30 numbered genes, which include a total of 7 cancer-relevant genes. (B) Input scenario for algorithm: a clinical genomics report noting amplification of three genes in this region. (C) Copy number inference for genes in regions bounded by reported amplified genes. (D) Copy number inference for genes surrounding regions bounded by reported amplified genes.\n\nInferAMP is accessed via a web browser and has been tested on commonly used browsers such as Chrome, Firefox, Internet Explorer and Microsoft Edge.\n\n\nUse cases\n\nThree use cases taken from genomic reports of patients at our clinic are presented:\n\nA case of esophageal adenocarcinoma with eight reported amplified genes (Table 1), which were resolved by InferAMP to four amplicons. The co-amplification of FGF3, FGF4 and FGF10 with CCND1 (which is likely the driver gene in this amplicon9) might indicate that consideration of FGFR inhibitors may not be helpful if these FGF genes are simply co-amplified passenger genes.\n\nA case of soft tissue sarcoma with five reported amplified genes (Table 2) which were resolved into three amplicons. In the absence of genomic context information, both PDGFRA and KIT might be considered as potentially druggable targets. The demonstration that these are likely co-amplified in a relatively small amplicon might provide further support to this hypothesis. Clinically, both targets are inhibited by imatinib, making joint targeting with a single agent feasible in this case.\n\nThe third example is a breast cancer case from our clinic with three reported amplified genes (Table 3). The report highlights one region on chromosome 5, and two regions on chromosome 7. The latter predicted amplicons share a nearby boundary at 7q22.3 suggesting the possibility that there is a regional amplification on 7q encompassing both sets of genes. In this case, MET was judged to be a possible driver amplicon, and the patient had a very strong response to a MET inhibitor10.\n\n\nDiscussion\n\nWe have described a straightforward tool to provide additional genomic context to aid interpretation of amplifications in somatic cancer sequencing reports. Use of this tool may aid decision-making by healthcare professionals about therapeutic options.\n\nThe method relies on the accuracy with which test vendors report gene amplification calls. In testing, we identified a small number of cases in which two amplicons were inferred in very close proximity (e.g. Use case 3), which raises the possibility that the assayed gene between the two regions is erroneously not called as amplified. In cases with two or more closely co-located amplicons, users should consider that there is a strong possibility of a regional amplification encompassing both predicted amplicons. Future assays with larger number of genes or more sensitive amplification calling algorithms will likely permit more accurate refining of the boundaries of individual amplicons.\n\nBecause the coverage across the genome is somewhat sparse, refining the amplicon boundaries is more challenging than with a more high-density approach like SNP arrays. The primary purpose is to list genes that are potentially co-amplified with a gene identified by a test vendor as possibly actionable in order to allow healthcare professionals to gain further insight into the likelihood that the listed gene is truly the driver gene in that amplicon. Accordingly, we do not distinguish in the report between genes that are likely co-amplified (red genes, Figure 1) from the boundary region genes which are possibly co-amplified (green genes, Figure 1). In any case in which a healthcare professional might consider targeting a non-assayed gene predicted by this algorithm to be amplified (e.g. LIFR11 in Use case 2 and Use case 3), further clinical testing to directly confirm gene amplification would be warranted.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\n\nSoftware availability\n\nSoftware available at: http://inferamp.org.\n\nSource code available from: https://github.com/paraickenny/inferAMP.\n\nArchived source code at time of publication: http://doi.org/10.5281/zenodo.324920112.\n\nLicense: MIT License.", "appendix": "Grant information\n\nThis project was supported by the Gundersen Medical Foundation. P.K. holds the Dr. Jon & Betty Kabara Endowed Chair in Precision Oncology.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nZack TI, Schumacher SE, Carter SL, et al.: Pan-cancer patterns of somatic copy number alteration. Nat Genet. 2013; 45(10): 1134–40. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nParakh S, Gan HK, Parslow AC, et al.: Evolution of anti-HER2 therapies for cancer treatment. Cancer Treat Rev. 2017; 59: 1–21. PubMed Abstract \| Publisher Full Text\n\nLim SH, Sun JM, Choi YL, et al.: Efficacy and safety of dovitinib in pretreated patients with advanced squamous non-small cell lung cancer with FGFR1 amplification: A single-arm, phase 2 study. Cancer. 2016; 122(19): 3024–31. PubMed Abstract \| Publisher Full Text\n\nVan Cutsem E, Bang YJ, Mansoor W, et al.: A randomized, open-label study of the efficacy and safety of AZD4547 monotherapy versus paclitaxel for the treatment of advanced gastric adenocarcinoma with FGFR2 polysomy or gene amplification. Ann Oncol. 2017; 28(6): 1316–24. PubMed Abstract \| Publisher Full Text\n\nSepúlveda-Sánchez JM, Vaz MA, Balañá C, et al.: Phase II trial of dacomitinib, a pan-human EGFR tyrosine kinase inhibitor, in recurrent glioblastoma patients with EGFR amplification. Neuro Oncol. 2017; 19(11): 1522–31. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHodi FS, Corless CL, Giobbie-Hurder A, et al.: Imatinib for melanomas harboring mutationally activated or amplified KIT arising on mucosal, acral, and chronically sun-damaged skin. J Clin Oncol. 2013; 31(26): 3182–90. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTan O, Shrestha R, Cunich M, et al.: Application of next-generation sequencing to improve cancer management: A review of the clinical effectiveness and cost-effectiveness. Clin Genet. 2018; 93(3): 533–44. PubMed Abstract \| Publisher Full Text\n\nTate JG, Bamford S, Jubb HC, et al.: COSMIC: the Catalogue Of Somatic Mutations In Cancer. Nucleic Acids Res. 2019; 47(D1): D941–D7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nQie S, Diehl JA: Cyclin D1, cancer progression, and opportunities in cancer treatment. J Mol Med (Berl). 2016; 94(12): 1313–26. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nParsons BM, Meier DR, Gurda GT, et al.: Exceptional Response to Crizotinib in an MET-Amplified Triple-Negative Breast Tumor. JCO Precis Oncol. 2017; 1: 1–6. Publisher Full Text\n\nHall BR, Cannon A, Thompson C, et al.: Utilizing cell line-derived organoids to evaluate the efficacy of a novel LIFR-inhibitor, EC359 in targeting pancreatic tumor stroma. Genes Cancer. 2019; 10(1–2): 1–10. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKenny P: paraickenny/inferAMP: inferAMP command line and web tool versions (Version v1.0.2). Zenodo. 2019. http://www.doi.org/10.5281/zenodo.3249201" }	[ { "id": "50604", "date": "11 Jul 2019", "name": "Andrew C. Nelson", "expertise": [ "Reviewer Expertise Molecular genetic pathology", "anatomic pathology", "breast cancer", "gynecologic cancer", "colorectal cancer", "tumor microenvironment", "cancer biology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this manuscript, Dr. Kenny describes a useful and straightforward web-based informatics tool that enables genetics professionals to rapidly and potentially more accurately infer genomic relationships of copy number alterations (CNA) disparately reported as individual genes in clinical genomics reports. This will aid both clinicians in working/teaching conferences (such as molecular tumor boards) and translational researchers reviewing archived clinical data.\n\nThe rationale for the tool and the overview of the method is clear. The web interface is easy to use and provides a clear result report; code is available at Github.\n\nI have several points for improvement, clarification, and commentary on the manuscript and the software tool:\n\nMajor Considerations\n\nI recommend that the assayed genes without amplification which are used by the algorithm as boundary genes be explicitly reported in the interface. In Fig 1D, these are represented as genes 4 and 27; as a minor point, I believe the legend for Fig 1D needs to be updated to be congruent with the figure. In the current interface report, the region of potential contiguous amplifications are only reported as cytobands. Specifically reporting the identity of these boundary genes would help genetics professionals to more precisely quality control the interpretation of both the original genomics report and the interface output.\n\nThe tool is currently static configured for FoundationOne (F1); it would be beneficial to configure the web interface to allow (perhaps through a drop down menu) selection of other reasonably common cancer genomics reports which offer CNA data such as: Caris, Tempus, and the Illumina products TST170 and TSO500 (which are being deployed by some academic laboratories for clinical testing). Ultimately, an option to input a text file with HGVS gene names would be beneficial.\n\nI would suggest an expanded introduction or discussion about the status of clinical utility for gene amplification from comprehensive genomic profiling NGS assays. For example, only ERBB2 amplification is specifically included in the FDA premarket approval of the F1 assay (https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpma/pma.cfm?id=p170019, accessed 07/10/2019). The most common diagnostic approach for gene amplification in clinical practice and clinical trials is FISH, frequently normalized to the centromeric copy number of the same chromosome on which the gene of interest resides (for example, see references 2-5 of the manuscript). There are, of course, significant pitfalls in the use of FISH as a longer established clinical standard, but I am not immediately aware of comprehensive accuracy assessments for somatic CNA analysis across broad numbers of genes by NGS, particularly in unpaired tests (i.e. no patient-specific normal sample for comparisons). A more in-depth review of these considerations and limitations in the introduction or discussion would equip readers to better interpret the expected output of datasets and any results generated using the tool.\nMinor Comments and Commentary\nPlease clarify in the methods if the COSMIC Cancer Gene Census list in the tool has been filtered for type of alteration (i.e. SNV and small indel vs. gene rearrangement vs. CNA). Or is it the full 719 CGC list as referenced?\n\nIt may be valuable to consider cross-referencing the cytobands and/or cancer-associated genes within the predicted CNA region against regions/genes commonly copy number altered in pan-cancer analyses. For example: Beroukhim et al.[ Nature 20101, Zack et al. Nature Genetics 20132, Ohshima et al. Scientific Reports 20173. Presenting this information in the report interface in future iterations would improve the quality and utility of the software tool.\n\nThe ability to specifically annotate case results against disease-specific databases (i.e. TCGA projects) would also be valuable (see below as well).\n\nCommentary on use cases.\n\nThe use cases provide a reasonable snapshot of how the tool functions and can be applied. Specifically, use case 1 highlights a common CNA in cancer (11q13.3) which has been described previously (Zack et al. Nature Genetics 20132) and this specific amplicon is commonly seen in clinical genomics reports at our institution. The question of utility around FGFR inhibitors has arisen in our own molecular tumor boards based on the co-amplified FGF ligand genes, when in reality CCND1 is most likely the significant driver.\n\nUse case 2 highlights potentially actionable genes (KIT, PDGFRA) located in the same cytoband (4q12). However, the region of potential co-amplification crosses the Chr4 centromere (Chr4p15.31-4q12). It may be beneficial to further discuss how carefully to interpret amplicons which span both chromosomal arms. In this particular case, I infer that no reported amplified genes are present on 4p; but no boundary genes were analyzed centromeric to the PDGFRA locus. This seems worth a deeper discussion in the text.\n\nUse case 3 highlights an important point about potential drop out of amplification calls in NGS data. Dr. Kenny proposes that the entire amplicon on 7q is amplified, which is a reasonable hypothesis. It is important to note that specific bioinformatics pipelines and wet-bench library prep methods will not have equivalent analytic sensitivity/specificity for amplification calls for every captured gene. Case 3 is also interesting because the amplicon included CDK6, which is a target for FDA approved drugs in hormone receptor positive metastatic breast cancer. It might be interesting for Dr. Kenny to further comment on whether CDK inhibitors were considered less likely based on the triple negative hormone receptor status (PMID: 300386704).\n\nI ran several of my own cases through InferAmp. Of interest, a case of high grade serous epithelial ovarian carcinoma is illustrative of the utility of the tool. This case had 8 separate gene amplifications reported by F1, including: KRAS, FGF23, FGF6, and CCND2 (all on Chr12p). The static clinical report indicated KRAS amplification was potentially relevant for MEK inhibitor therapy. Nevertheless, my molecular tumor board noted this amplification contig and noted that 12p is commonly amplified in high-grade serous carcinoma (TCGA); therefore it was unlikely to be a patient-specific driver alteration. The InferAmp quickly and accurately identified this amplicon, with the caveat that a potentially erroneous “break” was present at 12p13.1 (similar to Dr. Kenny’s use case 3).\n\nFinally, the final sentence of the manuscript cautioning that inferred co-amplified genes should be confirmed by CLIA-validated assays cannot be emphasized enough.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly", "responses": [ { "c_id": "4934", "date": "26 Sep 2019", "name": "Paraic Kenny", "role": "Author Response", "response": "I greatly appreciate the thoughtful and supportive feedback and suggestions. These have been very helpful in revising the manuscript.Major:1. We have added a new checkbox to the input page allowing users to request “verbose” output. In this case, the boundary genes are explicitly listed in the report.2. We have added gene lists corresponding to seven additional cancer genomic reports (Foundation CDx, Trusight, Tempus, Caris etc.). We have also added a text box into which users can paste custom lists of genes for other unsupported assays. Because the accuracy of amplicon inference is related to the density/distribution of assayed genes across the genome, we added a comment on this to reports generated using panels with low gene numbers.3. We have added a brief section to the discussion to discuss this, citing a couple of recent publications addressing the cross-comparison of NGS and FISH-based methodologies for CNA assessment.Minor:1. We used the 723 Tier 1 and 2 variants in the COSMIC gene census. This is now explicitly stated in the manuscript. While it may be reasonable to restrict the list to only those genes with known CNAs, we considered it best to use the broader list and trust the user to make the appropriate assessment. This avoids potentially missing situations where an oncogene that is typically activated by point mutation is activated by amplification under rare/unusual circumstances.2. This is an excellent suggestion. We evaluated a number of CNA databases/datasets but implementing a straightforward and automated method for importing the data to annotate inferred amplicons in our reports was not feasible.Use cases:The point about drop-outs leading to inference of two adjacent amplicons where, in fact, there is just one true amplicon which contains an assayed gene not reported as amplified is an important one. As we now report the boundary genes (See point 1), we have implemented an additional check so that if a report contains two inferred amplicons which share a single boundary gene, this is explicitly flagged so that the user will be asked to consider the possibility that the “boundary gene” may be erroneously reported as non-amplified." } ] }, { "id": "51604", "date": "15 Aug 2019", "name": "Oscar Krijgsman", "expertise": [ "Reviewer Expertise Bioinformatics", "DNA copynumber profiling" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe author describes a very straight-forward tool to infer possibly amplified genomic regions based on the amplification status of single genes from a targeted panel. InferAMP reports the genes that are likely co-amplified and in addition infers the size and the number of potentially co-amplified genes that were not assessed in the assay and outputs the COSMIC genes found in the genomic region. This tool is developed for oncologists to better understand the reported amplified genes in genomic context. The developed web-based tool is therefore important and will be the preferred way of using this tool for oncologists. It is also good to see the full code and data tables used are available on Github.\n\nInferAMP is a very easy tool to use and functions as advertised. The associated manuscript is well readable and explains the functionality satisfactory. However, after reading the manuscript and testing InferAMP I have a few questions that need to be answered.\n\nPoints to be addressed:\nThe web-based tool is now only suitable for FoundationOne assays. Although this is very useful for many oncologists not all institutes use this assay. I am aware that the the command-line version of the tool has the possibility to provide different genes set and therefore works with additional assays. However, a command-line version will not be suitable for oncologists. Additional functionality of the web-based would greatly improve the usability of InferAMP.\n\nThe output of InferAMP includes the genes that are possibly co-amplified and the genes in the amplicon that are mentioned in the COSMIC database. In addition, the results mention the potentially co-amplified genes (for example 214 genes when running CCND1, FGF3, FGF4 and FGF19, Use case 1). A list of these genes is currently not available. It would be useful to output these lists of genes in addition to the COSMIC as not all targetable genes will necessarily be in the COSMIC list.\n\nCurrently the COSMIC list is used to provide an additional rationale to prioritize genes and identify genes suitable for targeting. The rationale for using this list is not provided in the text. A little more background and explanation for choosing this list would be helpful. In addition, which COSMIC list is used exactly? COSMIC cancer census list1 or all genes mentioned in COSMIC?\n\nPoints for consideration but not necessary for this manuscript:\nFurthermore, I was wondering whether germline CNVs could affect the results of targeted panels, especially since patient-matched blood reference samples are not often used. If so, would it not be nice to add CNV data from for example the Database of Genomic Variants (http://dgv.tcag.ca)? This could identify amplifications that are not somatic but merely ‘normal’ differences between individuals (CNVs).\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4933", "date": "26 Sep 2019", "name": "Paraic Kenny", "role": "Author Response", "response": "I greatly appreciate the thoughtful and supportive feedback and suggestions. These have been very helpful in revising the manuscript.Major points:1. We have added gene lists corresponding to seven additional cancer genomic reports (Foundation CDx, Trusight, Tempus, Caris etc.). We have also added a text box into which users can paste custom lists of genes for other unsupported assays.2. We have added a new checkbox to the input page allowing users to request “verbose” output. If this is selected, all genes in the inferred amplicon are listed (i.e. not just the COSMIC genes).3. We used the full COSMIC cancer gene census list (723 Tier 1 and 2 genes). The goal is to indicate to users where other potentially cancer-relevant genes are likely co-amplified with the reported genes so that users may consider whether the reported gene is truly the cancer driver in that region.Minor Points:1. Germline CNVs are an interesting question, but beyond the scope of the current project. This is a problem best left to the developers/interpreters/reporters of the various genetic assays. Our goal has been to provide an easy to use tool with which to assess genes reported as amplified after having passed some QC by commercial testers or clinical labs." } ] } ]	2	https://f1000research.com/articles/8-807
https://f1000research.com/articles/8-1692/v1	26 Sep 19	{ "type": "Research Article", "title": "Developing a phosphodiesterase-5 inhibitor assay to detect counterfeit drugs containing phosphodiesterase-5 inhibitors using spectrophotometry", "authors": [ "Aziemah Azizi", "Asrin Tengah", "Mark I. R. Petalcorin", "Aziemah Azizi", "Asrin Tengah" ], "abstract": "Background: Phosphodiesterase-5 inhibitors (PDE5i) are the first line of drugs used for the treatment of erectile dysfunction. PDE5i inhibits the activities of phosphodiesterase-5 (PDE5) found in corpus cavernosum reducing production of guanosine monophosphate (GMP) that results in vasodilation of blood vessels thereby prolonging penile erection. Various counterfeit drugs adulterated with unknown levels of PDE5i pose a danger with reported undesirable adverse effects on patients. Methods: We developed an assay to detect counterfeit drugs containing PDE5i using spectrophotometry based on a colour change of malachite green from blue to green when exposed to inorganic phosphate (Pi) measured at 630 nm. Initially, a standard graph of a known Pi concentration was established ranging from 0 to 20 µM followed by a dual biochemical reaction system of converting GMP from cGMP by PDE5, generating Pi from the first reaction’s product by calf intestinal alkaline phosphatase (CIAP), and then measuring Pi using an equation derived from the generated standard curve. Dilutions of CX1A, a counterfeit drug sample adulterated with the PDE5i sildenafil, were prepared via ultrasonication and filtration for assay testing and evaluation. Results: Sil was detected from the two separate samples comprised of low and high dilutions of CX1A quantified as 0.673% and 0.407%, respectively, based on a standard curve of pure sildenafil established from 0.1 nM to 300 µM. Conclusions: This PDE5i assay that we developed has the potential to become an alternative analytical method for detecting PDE5i showing comparable results when evaluated using HPLC.", "keywords": [ "Phosphodiesterase-5 inhibitor", "Sildenafil", "counterfeit drugs", "PDEi assay" ], "content": "Introduction\n\nPhosphodiesterase-5 inhibitors (PDE5i) are synthetic drugs for clinical treatment of erectile dysfunction prescribed only by a qualified physician1. As PDE5 regulates penile erection by hydrolyzing cGMP, inhibiting the enzyme elevates cGMP level resulting in vasodilation and maintenance of penile blood flow. PDE5i are widely used in the illegal market as an adulterant in counterfeit drugs, especially in herbal supplements claiming the enhancement of male sexual vitality in a natural way2. The use of these counterfeit drugs with unknown content has posed significant health risks to patients. Hence, we developed an assay for easier detection and quantification of counterfeit drugs containing PDE5i using spectrophotometry.\n\nHerein, a known counterfeit herbal drug, CobraX (CX1A), was investigated to develop a potential analytical assay to detect PDE5i (i.e. sildenafil citrate) by utilizing the by-products (i.e. inorganic phosphate (Pi)) that were released from a reaction among the substrates comprised of cGMP, enzyme PDE5 and PDE5i (often adulterated in counterfeit drugs, i.e. CX1A). The colorimetric change in malachite green (MLG) assay was utilized to indirectly quantify the amount of Sil in counterfeit drugs through the detection of Pi released from the PDE5-mediated hydrolysed cGMP. The PDE5i assay that we developed in this study has been evaluated against HPLC and has shown high potential for laboratory and field kit applications to detect and measure the PDE5i content in any drug.\n\n\nMethods\n\nPotassium phosphate monobasic (KH2PO4), guanosine 3’,5’-cyclic monophosphate (cGMP), polyvinyl alcohol (PVA), sildenafil citrate salt, calf intestinal alkaline phosphatase (CIAP), human phosphodiesterase 5A1 (PDE5), ammonium molybdate tetrahydrate (MolyB) and malachite green oxalate free base (MLG) were all obtained from Sigma-Aldrich. The adulterated drug, CobraX (CX1A) was kindly donated by Brunei law enforcement agencies. BiotekTM ELx808TM absorbance microplate reader (Thermo Fisher Scientific) was used to read absorbance measured at 630 nm in 96-well microplate-based assays by a spectrophotometric method using MLG as previously described3 was adapted with modifications.\n\nThe MLG assay used to detect the counterfeit PDE5i drugs consisted of complex formation between phosphomolybdate (MolyB) with Pi, which was released from the PDE5-mediated hydrolysed cGMP, to produce MolyB-Pi complex, which was subsequently used to bind with MLG oxalate. Under acidic conditions, the bound MLG-MolyB-Pi produced colour changes from blue to shades of green depending on the concentration of Pi3, based on the observation that high concentrations of Pi resulted in high formation of MLG-MolyB-Pi complexes and high absorbance values.\n\nA functional PDE5i assay consisted of five steps. Figure 1 gives a schematic overview of the protocol.\n\nThis developed protocol can be used to detect other types of PDE5i in any drug by substituting the compound in the reaction assay using the specific PDE5i of interest and counterfeit drug.\n\nStep 1: Establishment of standard curve for Pi. A fundamental part of the assay was to establish the limit range of Pi concentrations for the MLG assay. Using n=2 replications, 230 µl of Pi solution (0, 2, 5, 7.5, 10, 12.5, 17.5 and 20 µM) was prepared in 0.5 ml microtubes and 75 µl of stop solution (70% perchloric acid) was added. The solutions were mixed by vortex and 140 µl was aliquoted into each designated well of 96-well microplate for MLG assay.\n\nStep 2: Dual biochemical reaction assays for PDE5-cGMP/GMP-CIAP (PDE5 reaction assay). The optimised PDE5 reaction assay was performed to obtain the total Pi generated per reaction, which was subsequently used to calculate the amount of sildenafil in counterfeit drug samples. For reactions replicated n=3 times, 20 µl of PDE5 solution (2000 unit/µg PDE5) and 0.0001 U/µl CIAP solution were added to a 0.5 ml microfuge tube containing 160 µl reaction buffer (20 mM Tris-HCL at pH 7, 10 mM MgCl2, 0.10 mM EDTA). The solutions were vortexed followed by addition of 30 µl of 15 mM cGMP and incubation for 30 min at 37˚C. After incubation, 50 µl of stop solution was added to terminate the reaction and vortexed followed by centrifugation at 2000 rpm for 5 min. The solutions were then aliquoted into their designated wells in the 96-well microplate with a total volume of 140 µl per well for the MLG assay.\n\nThe percentage yield of Pi was calculated using the following formula:\n\nPercentage yield of Pi (%) =No.ofmolesofgeneratedPiNo.ofmolesofexpectedPi×100\n\nStep 3: Establishment of standard curve for sildenafil. The effect of different concentrations of sildenafil (1, 3, 10, 30, 100 and 300 nM) upon exposure to PDE5 reaction was investigated and subsequently used to generate a standard curve for pure sildenafil with n=4 repeats. First, 40 µl of PDE5-CIAP mixture was added to a 190 µl of sildenafil-cGMP mixture in a 0.5 ml microtube followed by incubation at 37˚C for 30 min. After incubation, 50 µl of stop solution was added and mixed well using vortex followed by centrifugation at 2000 rpm for 5 min. The solutions were then aliquoted to their designated wells in a 96-well microplate with a total volume of 140 µl per well followed by MLG assay performance. The percentage of PDE5 activity of each concentration of sildenafil was calculated using the formula below:\n\nPercentage of PDE5 Activity (%) = =AbsorbanceofsildenafilAbsorbanceofoptimumPDE5reaction×100%\n\nStep 4: Quantification of sildenafil in CX1A samples via PDE5-CX1A reaction assay. A PDE5-CX1A reaction assay was performed to assess and quantify sildenafil in CX1A samples with n=2 repeats. To a 0.5 ml microtube containing 140 µl of reaction buffer, 20 µl of CX1A solution (0.1 mg/ml and 0.01 mg/ml) and 30 µl of 15 mM cGMP were added and mixed using vortex. For a reaction to start, 40 µl of PDE5-CIAP mixture was added followed by incubation at 37˚C for 30 min. After incubation, 50 µl of stop solution was added to the reaction and mixed well using vortex followed by centrifugation at 2000 rpm for 5 min. The solutions were then transferred to their designated wells in a 96-well microplate with a total volume of 140 µl per well followed by MLG assay.\n\nPercentage of PDE5 activity (%) =AbsorbanceofCX1AdrugsolutionAbsorbanceofoptimumPDE5reaction×100\n\nThe concentration of sildenafil was calculated using the following formula, which was derived from the standard curve of sildenafil:\n\nY = Bottom + Top−Bottom1+10(logIC50−X)slope\n\nWhere Y = percentage of PDE5 activity, X = Concentration of sildenafil, top value: 99.99, bottom value: -1.230, logIC50 (half-maximal inhibitory concentration): -0.505 and slope: -0.722\n\nStep 5: MLG assay. After the desired solutions were aliquoted into their designated well in the 96-well microplate, the MLG assay was performed, where 14 µl of MolyB reagent (50 mM MolyB in 3.4 M sulphuric acid) was added to each well containing the mixture, mixed thoroughly and incubated for 5 min. Then, 26 µl of MLG reagent (1.0 mM MLG, 6.0 mM sulphuric acid, 0.16% PVA) was added, mixed well and incubated for 10 min. After incubation, the absorbance was measured using BiotekTM ELx808TM absorbance microplate reader at 630 nm.\n\nStatistical comparisons were made by means of two-way repeated measures Analysis of Variance (ANOVA). The data were analysed with Prism software (Version 6, Graph Pad Inc.). The results were shown as mean ± S.E.M and n represents the number of repeats in each experiment. P values of 0.05 or lower was considered to be statistically significant.\n\n\nResults\n\nPi concentrations ranging from 0 µM to 20 µM were used to generate a standard curve for Pi. Table 1 and Figure 2 show the relative absorbance values for Pi (0 µM to 20 µM) and the linear relationship between absorbance and concentration of Pi, respectively. The equation derived from the standard curve was used to quantify the amount of Pi generated from PDE5 reaction. Raw results of this assay, alongside each of the assays performed in this study, are available as Underlying data4.\n\nAU, arbitrary units.\n\nThe correlation allows the establishment of standard curve of Pi.\n\nUsing an optimized PDE5 reaction, 1.842 moles of Pi was generated as shown in Table 2.\n\n*Calculated using an equation derived from standard graph of Pi.\n\nTable 3 shows the percentage of PDE5 activity at each concentration of sildenafil. Using the values obtained, a sigmoidal curve was established between PDE5 activity and log concentration of sildenafil, with an IC50 of 0.3124 µM (Figure 3).\n\nSigmoidal relationship was established for standard curve of pure Sil.\n\nFrom the PDE5-CX1A reaction assay, the percentage values of PDE5 activity using low (0.1 mg/ml) and high dilution (0.01 mg/ml) samples of CX1A were 31.33% and 78%, respectively. Using the equation derived from the sigmoidal curve of sildenafil, the concentration of sildenafil was determined as the percentage of sildenafil in low and high dilutions of CX1A samples, i.e. 0.673% and 0.407%, respectively. The mass of sildenafil in CX1A tablet is 3.149 mg. The calculated values were as tabulated in Table 4 and Table 5.\n\nPDE5, phosphodiesterase-5.\n\n\nDiscussions and conclusion\n\nColour changes from blue to shades of green were observed with increasing Pi concentrations. This indicates that a complex of Pi with MolyB, and subsequently MLG, was formed. Low concentrations of Pi resulted in less formation of the Pi-MolyB-MLG complex, hence the observation of a lighter coloration with lower absorbance values. Increasing Pi concentrations led to greater binding and higher formation of complexes, indicated by high absorbance values. However, the linearity between concentration and absorbance was only observed up to 20 µM, indicating that beyond this maximum concentration range, the coloration of MLG with Pi-MolyB complexes was completely saturated. The Pi concentrations from 0 µM to 20 µM were used to develop the standard curve for Pi, which was needed in the later stages to quantify the amount of Pi released from PDE5 reaction. An important step to validate the reliability of this method is the linear correlation between sample concentration and the amount of activity measured. Reproducible satisfying linear correlations were observed using 0 µM to 20 µM Pi, with line of regression of 0.993 ± 0.001 (n=2) demonstrating that the established standard curve for Pi was robust and reliable to measure Pi.\n\nResults of several preliminary studies using different types of buffer, working pH and types of substrate showed that a reaction buffer with pH 8 and cGMP were the most suitable conditions for the PDE5 reactions. The reaction buffer consisting of Tris-HCl, EDTA affected the activity of CIAP and MgCl2, provided cGMP-dependent Mg2+ that further supported the catalytic activity of PDE5. Although pH 8 favored CIAP’s activity more than PDE5 limiting GMP production, it allowed both dual reactions to occur and produce better results. Based on specificity of cGMP to PDE5, cGMP was used as the substrate.\n\nApproximately 100 units (a conversion of 1 picomole of cGMP to GMP per minute is defined as one unit) of PDE5 with a 30 min incubation was used for each PDE5 reaction. To completely allow all generated GMP to release Pi, about 10-fold of 0.001 unit CIAP concentration (where conversion of 1 µM of GMP to Pi per minute is defined as one unit) was used with 30 min incubation time. Therefore, 3 nmol of both cGMP and Pi production was expected per PDE5 reaction. However, the observed efficiency of the reaction was only 60% of the expected amount (1.842 moles), suggesting that the reaction conditions, although operational, did not yet achieved the optimal level for the dual biochemical reaction assay. This is despite the fact that results were within the acceptable range with more than half of the expected cGMP-GMP-Pi conversion taking place.\n\nSildenafil was used as the sole reference for PDE5i in this study. Having similar structures, sildenafil competes with cGMP for the access to the catalytic site and is therefore known as a ‘competitive inhibitor’. Sildenafil does not interact with the allosteric cGMP-binding site in PDE55. Among the members of PDE enzyme families, sildenafil is highly selective towards PDE5, with a high inhibitory affinity for the catalytic site, as compared with the ~600 to 25,000-fold lower affinity of the natural substrate, cGMP6. In general, PDE5 activity decreases as sildenafil concentration increases. At lower concentrations of sildenafil, the amount of cGMP surpasses sildenafil, increasing the chance of cGMP binding to PDE5, thus allowing greater PDE5 activity. However, as the concentration of sildenafil increases, it displaces cGMP and binds to PDE5, causing a decrease in PDE5 activity. The total loss of PDE5 activity would occur when sildenafil completely inhibits all PDE5 molecules.\n\nA sigmoidal relationship was established between PDE5 activity and different concentration of sildenafil, demonstrating positive cooperative binding. This means that the binding of the enzyme with its substrate at one binding site affects the affinity of other sites for their substrates, allowing a rapid and increased coordinated enzymatic activity. This is consistent with a PDE5 profile that has both catalytic and allosteric binding sites for its substrates. However, since sildenafil is considered a competitive inhibitor, which binds at the catalytic site of PDE5, the sildenafil inhibitory profile shows negative cooperative binding (represented by negative hill slope value) presumably reflecting the changes in kinetic characteristics caused by cGMP-induced PDE5 activation7. The IC50 of sildenafil towards PDE5 in this study of 312.4 nM was about ~40-70-fold higher than previous studies that showed IC50 of around 3.5–8 nM8. This variation could be attributed to the dependency of IC50 values on several factors such as cGMP concentration, the source and extraction method of enzymes, the reaction condition, the number of samples and other factors affecting the experimental design9.\n\nThe amount of sildenafil in CX1A samples determined in this study was validated using HPLC producing comparable results (data not shown). The low and high dilutions of CX1A resulted in 0.673% and 0.407% of the calculated percentage of sildenafil in CX1A, respectively. Other types of PDE5i and counterfeit drugs could also be tested and investigated using the PDE5i assay and employing optimal conditions and appropriate controls. In conclusion, the PDE5i assay developed in this study has the potential to become an alternative method for the detection of counterfeit drugs containing PDE5i.\n\n\nData availability\n\nOpen Science Framework: PDE5i Assay. https://doi.org/10.17605/OSF.IO/PV59F4.\n\nThis project contains ‘Aziemah et al. 2019 Raw_data for developed PDE5i assay.xlsx’, which includes raw data for all experiments performed in this study.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).", "appendix": "Grant information\n\nThis work was funded under the University Research Grant of the Universiti Brunei Darussalam to M.I.R.P.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nHuang SA, Lie JD: Phosphodiesterase-5 (PDE5) Inhibitors In the Management of Erectile Dysfunction. P T. 2013; 38(7): 407–419. PubMed Abstract \| Free Full Text\n\nLebel P, Gagnon J, Furtos A, et al.: A rapid, quantitative liquid chromatography-mass spectrometry screening method for 71 active and 11 natural erectile dysfunction ingredients present in potentially adulterated or counterfeit products. J Chromatogr A. 2014; 1343: 143–151. PubMed Abstract \| Publisher Full Text\n\nZhu S, Gan Z, Li Z, et al.: The measurement of cyclic nucleotide phosphodiesterase 4 activities via the quantification of inorganic phosphate with malachite green. Anal Chim Acta. 2009; 636(1): 105–110. PubMed Abstract \| Publisher Full Text\n\nPetalcorin M, Azizi A, Tengah A: PDE5i Assay. 2019. http://www.doi.org/10.17605/OSF.IO/PV59F\n\nCorbin JD, Francis SH: Cyclic GMP phosphodiesterase-5: target of sildenafil. J Biol Chem. 1999; 274(20): 13729–13732. PubMed Abstract \| Publisher Full Text\n\nFrancis SH, Morris GZ, Corbin JD: Molecular mechanisms that could contribute to prolonged effectiveness of PDE5 inhibitors to improve erectile function. Int J Impot Res. 2008; 20(4): 333–342. PubMed Abstract \| Publisher Full Text\n\nRybalkin SD, Rybalkina IG, Shimizu-Albergine M, et al.: PDE5 is converted to an activated state upon cGMP binding to the GAF A domain. EMBO J. 2003; 22(3): 469–478. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSaenz de Tejada I, Angulo J, Cuevas P, et al.: The phosphodiesterase inhibitory selectivity and the in vitro and in vivo potency of the new PDE5 inhibitor vardenafil. Int J Impot Res. 2001; 13(5): 282–290. PubMed Abstract \| Publisher Full Text\n\nGresser U, Gleiter CH: Erectile dysfunction: comparison of efficacy and side effects of the PDE-5 inhibitors sildenafil, vardenafil and tadalafil--review of the literature. Eur J Med Res. 2002; 7(10): 435–446. PubMed Abstract" }	[ { "id": "54386", "date": "30 Sep 2019", "name": "Dirk W. Lachenmeier", "expertise": [ "Reviewer Expertise analytical chemistry", "food chemistry" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors describe an application for spectrophotometric analysis of PDE5 inhibitors in counterfeit drugs or food supplements. The application could be interesting in settings where more advanced analysis such as HPLC or LC/MS is unavailable.\nMajor deficits:\nNo validation data at all is presented besides linearity (e.g. detection limits, accuracy, precision, selectivity)\n\nThe discussion section does not provide a comprehensive discussion at all. The results should be placed into the state of the art, i.e. compared to other methods such as HPLC or NMR. Furthermore, the advantages and limitations of the current research must be discussed. A major limitation is, for example, the lack of a real applicability study apart from one single sample. It is not demonstrated that the method works “in the real world” with all problems regarding matrix interferences or interferences from other similar compounds or isomers. For example, the RASFF system of the EU reports that currently sildenafil thiono analogues have a high incidence in illegal food supplement (i.e. hydroxythiohomosildenafil, hydroxyhomosildenafil and thiosildenafil)[reference],[reference]. Would such compounds be detected by your assay?\nThe following revisions are required:\nThroughout: There is a constant mismatch in units, which makes the paper difficult to follow. % should be avoided at all cost and replaced by meaningful units such as g/100 g, while µM could be recalculated to mg/L.\nThroughout: I dislike the non-chemical abbreviation Pi for inorganic phosphate. If the ionic state is known, e.g. PO42-, this should rather be stated.\nThroughout: please only present significant decimals in numbers considering validation results\nTitle: The title is rather long. Can the two times repetition of PDE5I be avoided perhaps?\nAbstract, conclusion: The data about “comparable results with HPLC” is not in the paper and should be deleted (or shown in the paper).\nIntroduction, 1st paragraph, 2nd last line: “easier detection”. Why is it easier? Compared to what it is easier? The sample preparation appears to be tiresome and could be more complicated than the sample preparation (dilution) required for HPLC or NMR analysis (e.g. see Monakhova Y et al., 20121).\nPage 3, right column, 2nd paragraph: HCL should read HCl\nTable 2: mole should read “mol” (or M) but see comment on units above?\nTable 4: what is the difference between final and actual concentration? Add footnotes\nTable 5: can the final result be reported as mg/kg?\nPage 7: “data no shown” for HPLC: why?\nFigure 1: explain abbreviation rxn\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [] }, { "id": "54383", "date": "01 Oct 2019", "name": "Adi Idris", "expertise": [ "Reviewer Expertise Innate immunity" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nWork presented by Azizi A et al described a spectrophotometric assay to detect the presence of phosphodiesterase-5 inhibitors. This could used as a basis in the early stages in a pipeline to develop a method to detect PDE5i in suspect counterfeit drugs. Though preliminary and brief in nature, these findings shed light on the potential use of this assay from an application standpoint. However, the authors are asked to address some comments below before this manuscript is fully acceptable:\nGrammatical and language mistakes found throughout. Recommend English proofreading\n\nAuthors are asked to define CX1A in the abstract\n\nThe term \"Sil\" was used in the abstract and introduction. Please define at first use or change to Sildenafil.\n\nTable 1 is unnecessary as Fig 2 clearly depicts this more clearly. This also applies to Table 3. If need be put the raw data as supplementary data\n\nThe results section was poorly written. It is too short and reads like a technical document. It lacks thought and short conclusive statements. Either re-write this section or consider merging the results section with the discussion section (if allowed by journal guidelines)\n\nTable 4 and 5 should be merged into one\n\nThe beginning of the discussion section reads poorly, making it hard for readers to grasp the essence of the paper. Authors should firstly highlight the main findings from their study then critically analyse their data.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1692
https://f1000research.com/articles/8-1685/v1	24 Sep 19	{ "type": "Research Article", "title": "Vaccine impact on antimicrobial resistance to inform Gavi, the Vaccine Alliance’s 2018 Vaccine Investment Strategy: report from an expert survey", "authors": [ "Maya Malarski", "Mateusz Hasso-Agopsowicz", "Adam Soble", "Wilson Mok", "Sophie Mathewson", "Johan Vekemans", "Maya Malarski", "Mateusz Hasso-Agopsowicz", "Adam Soble", "Wilson Mok", "Sophie Mathewson" ], "abstract": "Background: While the rise of antimicrobial resistance (AMR) has been recognised as a major public health problem, the value of vaccines to control AMR is poorly defined. This expert survey was launched with the aim of informing the 2018 Vaccine Investment Strategy through which Gavi, the Vaccine Alliance prioritises future vaccine funding. This exercise focused on both vaccines currently supported by Gavi and under consideration for future funding. Methods: The relative importance of pre-defined criteria as drivers of overall value of vaccines as a tool/ intervention to control AMR was assessed by 18 experts: prevention of mortality and morbidity due to resistant pathogens, antibiotic use prevented, societal impact, ethical importance and sense of urgency. For each vaccine, experts attributed scores reflecting the estimated value for each criterion, and overall value relative to AMR was derived from the value assigned to each criterion and their relative importance for each vaccine. Results: Mortality, morbidity due to targeted resistant pathogens, and antibiotic use prevented were considered the most important determinants of overall value. Pneumococcal, typhoid and malaria vaccines were assigned highest value relative to antimicrobial resistance. Intermediate value was estimated for specific rotavirus, cholera, respiratory syncytial virus (RSV), influenza, dengue, measles, meningitis and Haemophilus influenza type b- (Hib-) containing pentavalent vaccines. Lowest value relative to AMR was estimated for Japanese encephalitis, hepatitis A, yellow fever, rabies and human papilloma virus vaccine. Conclusions: In the future, more evidence-based, data-driven, robust methodologies should be developed to guide coordinated, rational decision making on priority actions aimed at strengthening the use of vaccines against AMR.", "keywords": [ "Vaccines", "antimicrobial resistance" ], "content": "Abbreviations\n\nAntimicrobial resistance (AMR) ; respiratory syncytial virus (RSV) ; Haemophilus influenza type b (Hib) ; World Health Organization (WHO) ; pneumococcal conjugate vaccine (PCV) ; inactivated polio vaccine (IPV) ; Vaccine Investment Strategy (VIS) ; diphtheria-tetanus-pertussis (DTP) ; typhoid conjugate vaccine (TCV) ; World Organisation for Animal Health (OIE).\n\n\nHighlights\n\nVaccines contribute to combat the growing threat of antimicrobial resistance\n\nA survey was undertaken to assess expert opinion about the value relative to antimicrobial resistance of potential vaccine investments considered by Gavi, the Vaccine Alliance\n\nEach expert assigned vaccine value by assigning a score to a set of pre-defined criteria, weighted for their relative importance: prevention of mortality and morbidity due to resistant pathogens, antibiotic use prevented, societal impact, ethical importance and sense of urgency\n\nExperts considered mortality, morbidity due to targeted resistant pathogens, and antibiotic use prevented as the most important determinants of overall value\n\nPneumococcal, typhoid and malaria vaccines were assigned highest value relative to antimicrobial resistance\n\nThis exercise will help shaping future evidence-based assessment of the public value of vaccines to contribute to control of antimicrobial resistance\n\n\nBackground\n\nAntimicrobial resistance (AMR) constitutes a major global health threat. Each year, an estimated 700,000 deaths result from infections with pathogens resistant to antimicrobial drugs, and this toll is expected to rise to 10 million by 20501. There is strong political momentum around the need to prioritise prevention of AMR pathogens in the global health agenda, as highlighted by the high-level meeting of the United Nations General Assembly on antimicrobial resistance2 and the Global Action Plan on Antimicrobial Resistance3. Recently, the World Health Organization (WHO) issued a list of priority pathogens against which new antibiotics should be developed4. With the expanding burden related to AMR, diminishing treatment options for a number of bacterial diseases, and a small pipeline of potential new therapeutics5, vaccines are increasingly recognised as an important complementary tool in controlling AMR6–10.\n\nVaccines have the potential to impact antibiotic resistance in several ways. Vaccines targeting a bacterial pathogen can reduce the vaccinated individual’s risk of an infection, not only protecting that individual but also possibly preventing further transmission of potential resistant strains. Both bacterial and viral vaccines can reduce the incidence of illnesses that prompt antibiotic use. This helps to reduce the selective pressure driving emergence of resistance on the targeted pathogen as well as the whole microbiome of the host. Some vaccines can also reduce bacterial carriage and the size of the pathogen population in the host, thereby reducing the risk of emergence of resistance and spread11. By decreasing the incidence of vaccine-preventable diseases, vaccines reduce care-seeking behaviour, such as attendance in health facilities, and thereby exposure to AMR pathogens12.\n\nEvidence which demonstrates the impact of some vaccines against AMR is available10. Within five years of the first introduction of pneumococcal conjugate vaccine (PCV) in the United States, prevalence of pneumococcal multidrug-resistant strains reduced by 57%, and the incidence of multidrug-resistant invasive pneumococcal disease decreased by 84% in children under two years old13. A similar post-introduction study in South Africa found that the incidence of invasive pneumococcal disease caused by PCV7 serotypes decreased by 85% in children not infected with human immunodeficiency virus14. A study estimated that global and widespread use of PCV could reduce the amount of antibiotics used for pneumonia patients by 47%, the equivalent of 11.4 million antibiotics days globally5. Similarly, introduction of Haemophilus influenzae type b (Hib) vaccines in the 1980s led to demonstrated reductions in the prevalence of Hib drug-resistant strains10,16. In Canada, an ecological study suggested that universal, free introduction of seasonal influenza vaccination in certain regions of the country led to a significant reduction of antibiotic prescriptions17. This remains a limited body of evidence, and the need for more research and modelling evaluation of vaccine impact on AMR has been clearly expressed18.\n\nGavi, the Vaccine Alliance was established in 2000 with the goal of creating equal access to new and underused vaccines for those living in lower-income countries. Since its inception, the Alliance has played a critical role in ensuring access to these vaccines, currently supporting vaccines that protect against 16 pathogens and contributing to the immunisation of more than 700 million children19. Every five years Gavi reviews its Vaccine Investment Strategy (VIS), to identify new vaccines and other immunisation products of most importance to Gavi-supported countries. The VIS sets new priorities for Gavi’s vaccine support programmes through an in-depth analysis of impact, cost, value and programmatic feasibility. Prioritised investments are included in Gavi’s portfolio. In 2017, Gavi commenced the development of its latest VIS, covering the 2021–2025 strategic period. Under consideration were 12 vaccine candidates and other immunisation products for endemic diseases, support for pandemic influenza preparedness, and inactivated polio vaccine (IPV) support post-202020.\n\nTo inform the prioritisation of Gavi’s potential future investments, an evaluation framework was developed. Various criteria including health impact, value for money, equity and social protection impact, economic impact and global health security were considered, along with other secondary criteria. This was the first time an indicator on “Impact of vaccination on AMR” was included in the evaluation framework for prioritisation of vaccine investments. The indicator sits within the broader “Global Health Security” criterion, which also separately considers the impact of vaccination on diseases with epidemic potential20. This article reports the results of an expert consultation that was used to assess the comparative value of potential vaccines against AMR, as one component of the wider prioritisation process.\n\n\nMethods\n\nIn view of the limited evidence available and time constraints to develop a more comprehensive analysis method for contribution to the VIS 2018, an approach based on expert opinion was selected through a collaboration between the Gavi Secretariat and the WHO Initiative for Vaccine Research. Both the current portfolio of Gavi-supported vaccines and those selected as part of the VIS 2018 exercise were considered, as detailed in Table 1. Previous Gavi investment prioritisation decisions did not formally consider the potential contribution of vaccines to control AMR.\n\nDTP (Diphtheria-, pertussis and tetanus); Penta (pentavalent vaccine); Td (tetanus and diphtheria booster vaccine); Meningitis ACWY or ACWXY (meningitis vaccine against A, C, W, Y strains or A, C, W, X, Y strains); RSV (respiratory syncytial virus); mAb (monoclonal antibody).\n\nBased on internal, institutional discussions, the list of criteria used to assess the value of vaccines against AMR was generated. Six criteria were identified, as presented in Table 2.\n\nExperts were identified from high impact AMR related research and publications; from the list of participants of AMR related meetings; and through a consultation with other experts and funders. They were eligible to participate in the study if they had a relevant expertise in AMR across the fields of infectious diseases, clinical microbiology, vaccinology, public health, epidemiology. Global geographic representation and gender balance were sought. Participants were included in the survey if they provided a complete set of responses. Background information (available as the extended data32) provided to each expert included a basic description about the vaccine investments considered and global disease burden data21.\n\nOn the 1st of February 2018 experts received an email and an Excel file with an introduction and background information tabs that explained the process of scoring each vaccine and pathogen (extended data32) and were asked to respond by the 13th of February. The file explained that experts are expected to weigh the relative importance of each criterion, by distributing 100% points across the six criteria. The experts were then asked to assign a score between 1 and 10 (where 1 = least effect of vaccine in future and 10 = greatest effect of vaccine in future) to each criterion, for all vaccines included in the assessment.\n\nThe average weight per criterion was calculated in Excel and expressed using lower and upper quartile ranges, minimum and maximum values, means and medians, and outliers defined as over 1.5interquartile range from the 1st or 3rd quartile. The vaccine scores (vaccine impact on AMR) for each criterion were calculated in Excel and expressed using average scores. The overall impact of vaccines on AMR (aggregate vaccine scores for all criteria) was calculated in Excel by multiplying the average weight per criterion and the vaccine scores (vaccine impact on AMR) assigned by each expert to different criteria. It was expressed using lower and upper quartile ranges, minimum and maximum values, means and medians, and outliers defined as over 1.5interquartile range from the 1st or 3rd quartile.\n\n\nResults\n\nThe survey was launched in February 2018. Of the 26 experts approached to participate, 18 completed the survey and all responses were complete. Experts were from the following countries: Australia, India, Republic of Korea, Republic of South Africa, Spain, Sweden, Switzerland, the United Kingdom and the United States. Ten experts were from research institutions, four from international organisations, two from public health institutes and two were independent (see underlying data33).\n\nFigure 1 shows the distribution of weights assigned to each of the criteria by the experts, out of a total of 100%. Higher importance was attributed to mortality (median 20%, interquartile range [IQR] 16-29%), antibiotic use prevented by the vaccine (20%, IQR 18-24%) and morbidity (18%, IQR 15-23%). Less importance was attributed to the sense of urgency due to AMR threat (15%, IQR 10-19%), societal impact (15%, IQR 10-16%) and ethical importance (5%, IQR 5-14%).\n\nLower error bar= minimum value; lower bound of the box= lower quartile range; black line inside the box=median; upper bound of the box=upper quartile range; upper error bar= maximum value; circle=mean; square = outlier (defined as over 1.5interquartile range from the 1st or 3rd quartile).\n\nFigure 2 presents the overall expert-estimated impact on AMR for the considered vaccine investments. Within the current Gavi portfolio, the three investments that had the highest estimated impact on AMR were those related to pneumococcal conjugate (median impact 7.00, IQR 5.71–8.00), typhoid conjugate (6.60, IQR 5.95–7.73) and pentavalent (diphtheria, pertussis, tetanus, hepatitis B and Hib), (4.10, IQR 3.59–5.47) vaccines. Among the potential vaccine investments considered as part of the VIS 2018 exercise, the three vaccines with the highest predicted impact on AMR were vaccines for malaria (6.85, IQR 5.93–7.46), cholera (4.15, IQR 2.43–5.40) and meningitis ACWY(X), (3.75, IQR 2.90–5.33). In contrast, the three vaccines currently funded by Gavi that received the lowest score for estimated impact on AMR were vaccines for Japanese encephalitis (1.30, IQR 1.00–1.86), yellow fever (1.00, IQR 0.95–1.40) and human papillomavirus (HPV; 1.00, 0.59–1.00). Among the vaccine investments considered as part of the VIS 2018 exercise, those with the lowest estimated AMR impact were vaccines for rabies (1.00, IQR 0.73–1.40), hepatitis A (1.00, IQR 0.33–1.20) and rabies immunoglobulin (1.00, IQR 0.73–1.00).\n\n1= least effect of vaccine and 10 = greatest effect of vaccine on AMR Lower error bar= minimum value; lower bound of the box= lower quartile range; black line inside the box=median; upper bound of the box=upper quartile range; upper error bar= maximum value; circle=mean; square = outlier (defined as over 1.5interquartile range from the 1st or 3rd quartile) Blue shading = Gavi portfolio vaccines; Green shading = VIS candidate vaccines PCV – pneumococcal conjugate vaccine; Pentavalent vaccine (diphtheria, pertussis, tetanus, hepatitis B and Haemophilus influenzae type B); DTP – diphtheria, tetanus & pertussis; RSV – Respiratory syncytial virus; mAb - monoclonal antibodies; Hep B – Hepatitis B; JE – Japanese encephalitis; HPV – Human papillomavirus; Hep A – Hepatitis A; RIG – Rabies immunoglobulin.\n\nMore detailed scores for each criterion are presented in Figure 3. Across the criteria, similar trends were observed in terms of how experts perceived specific vaccines, suggesting a lack of independence between the criteria. Pneumococcal conjugate, malaria and typhoid vaccines were consistently attributed the highest impact scores across all criteria. Intermediate impact across most criteria were estimated for rotavirus, cholera, influenza, dengue, measles, meningitis, pentavalent vaccines and for diphtheria-tetanus-pertussis- (DTP-) containing vaccine boosters, as well as vaccines and monoclonal antibodies against RSV. Across criteria, lower values were attributed to Japanese encephalitis, hepatitis, yellow fever, rabies and HPV-related vaccines.\n\n(where 1= least effect of vaccine in future and 10 = greatest effect of vaccine in future) Blue shading = Gavi portfolio vaccines; Green shading = VIS candidate vaccines; numbers above the bars indicate mean criterion scores for each vaccine. PCV – pneumococcal conjugate vaccine; Pentavalent vaccine (diphtheria, pertussis, tetanus, hepatitis B and Haemophilus influenzae type B); DTP – diphtheria, tetanus & pertussis; RSV – Respiratory syncytial virus; mAb - monoclonal antibodies; Hep B – Hepatitis B; JE – Japanese encephalitis; HPV – Human papillomavirus; Hep A – Hepatitis A; RIG – Rabies immunoglobulin.\n\nAll results and anonymised participants information are available as underlying data.\n\n\nDiscussion\n\nWhile the fight against AMR has been identified as a major public health priority, the value of vaccines in contributing to the control of AMR has been difficult to articulate7–9. Vaccines can contribute to the control of AMR through various, complex mechanisms, and the estimation of the full value of investments for new interventions requires defining value across multiple preference metrics22.\n\nThe results from this expert survey highlight the importance of investments in pneumococcal conjugate, malaria and typhoid vaccines, relative to other investments considered by Gavi, when their impact on AMR infections is taken into account.\n\nOf these, Gavi developed an Advanced Market Commitment for pneumococcal conjugate vaccine (PCV) in 2009. To date 60 countries have introduced the vaccine into their national schedules with support from Gavi. An unpublished evaluation from the Gavi secretariat in collaboration with John Hopkins University assessed that past Gavi support has averted 14 million antibiotic doses for pneumonia between 2011–15 alone. For 2016–2020, Gavi has committed to vaccinate hundreds of millions of children with vaccines against meningitis and pneumonia, estimated to potentially prevent over 100 million further antibiotic doses (personal communication Hope Johnson). The evaluation indicates that the perceived impact of PCV on AMR is related primarily to the vaccine’s demonstrated potential to reduce antibiotic consumption. This supports continued efforts to increase coverage and global use of PCV.\n\nGavi funding for typhoid conjugate vaccine (TCV) was made available at the end of 2017 with the first introductions planned for 2019, including in Zimbabwe. Pakistan carried out a campaign with Vi-polysaccharide typhoid vaccine in 2017 in response to a growing number of extensively drug resistant cases of Salmonella typhi. To date, 118,000 children aged between 6 months and 10 years have been vaccinated23. The specific role of TCV in the containment of resistant typhoid strains, and their utility in response to outbreaks with resistant strains should be further evaluated24.\n\nGavi is co-funding the RTS,S/AS01 malaria vaccine implementation pilot evaluation in Ghana, Kenya and Malawi25. Artemisinin-based combination therapies are recommended by WHO to treat uncomplicated Plasmodium falciparum malaria. Artemisinin-resistant malaria strains are present in South-East Asia. There is significant potential for spread, putting present malaria control strategies at risk26. The possible future wide-scale implementation of the malaria vaccine RTS,S/AS01 in young African children is unlikely to play a major role in the containment of artemisinin-resistant malaria strains initially, but the contribution to reduction in antibiotic use associated to febrile episodes may be significant27. Investigators are considering vaccine use in South-East Asia for the specific purpose of containing the spread of artemisinin resistance28.\n\nAs could be expected, vaccination strategies targeting bacterial pathogens generally received high scores for impact on AMR-related mortality, morbidity and sense of urgency. Experts also attributed significant value to some vaccines targeting viral pathogens when considering reduction in antibiotic use. Altogether investments in vaccines protecting against rotavirus, cholera, RSV, influenza, dengue, measles and meningitis, as well as the Hib-containing pentavalent vaccine were also considered to be of value, – more so than those related to Japanese encephalitis, hepatitis, yellow fever, rabies and HPV vaccines.\n\nOur results are broadly consistent with previous assessments of the value of vaccines in preventing AMR infections, including during a meeting held by the Chatham House in March 20177,8. The meeting identified several gaps, including developing an action framework for the use of vaccines to help contain AMR, gathering evidence and promoting research and development (R&D) of vaccines that could reduce antibiotic use and that protect against pathogens which are considered to be driving AMR threats. A poll was taken at the meeting to prioritise vaccines by their impact on AMR. Meeting participants identified tuberculosis, typhoid and influenza as the top-three priority vaccines for AMR impact8.\n\nA recent analysis conducted by Wellcome and the Boston Consulting Group investigated the role of vaccines in reducing AMR, targeting pathogens included in the WHO pathogen priority list of antibiotic-resistant bacteria4 as well as Mycobacterium tuberculosis. Health impact, probability of R&D success, and probability of uptake were considered for 18 priority pathogen groups. The report emphasised the need to increase the uptake of vaccines currently in use (PCV, Salmonella typhi and Hib), and accelerate market entry of vaccines for shigella, non-typhoidal salmonella, and Escherichia Coli29.\n\nLastly, the World Organisation for Animal Health (OIE) has conducted a review on prioritisation of diseases for which vaccines could reduce antimicrobial use in cattle, sheep and goats30. Taking into consideration the antibiotic use associated with diseases and current vaccine availability, the OIE orders vaccine research priorities for a number of vaccines targeting specific animal pathogens30.\n\nSeveral limitations of this value attribution exercise need to be considered. It was launched to contribute to the VIS 2018 exercise with significant time constraints. To address the lack of data on AMR, and data on vaccine impact on AMR pathogens, we decided to seek an expert opinion to complete this exercise. A total of 18 experts completed the survey, but a higher number would have been desirable. Typically, for multi-criteria decision analyses, criteria should be complete, non-redundant, non-overlapping and independent31. Some of these requirements were probably not completely met, and it is likely that mortality and morbidity outcomes influenced the expert assessment of sense of urgency, ethical importance and societal impact. The general homogeneity in the way specific vaccine investment were valued across different criteria suggests a lack of independence between criteria. It is also the case that some important considerations were not included in the value attribution framework, such as the availability of alternative interventions to control the pathogens considered. Despite its drawbacks, the exercise should be viewed as a preliminary attempt to estimate the comparative value of a range of vaccines in terms of their contribution to AMR control, and as a relevant benchmark for developing improved value attribution methodologies.\n\nIn the future, a more evidence-based, data-driven, robust methodology should be developed to assess the value of vaccines as a tool to control AMR, building on this exercise as well as the work conducted by Wellcome and Chatham House8,29. Complex modelling methods will likely have to be developed22. A long-term perspective should be adopted, as the public health need is likely to continue to be important. Confronted by data gaps, error margins are predicted to be large. However, as there is growing impetus for research in this area, more evidence will emerge and there will be opportunities to refine estimates. Future exercises should consider an expanded list of pathogens including tuberculosis, enteric pathogens and group A and B streptococcus, as well as other vaccines that were not part of the Gavi VIS 2018.\n\nThere is a need to define and align on priority actions to strengthen the use of vaccines to control AMR and coordinate their implementation. This could include the expression of priorities in evidence generation, such as advocating collection and analysis of antibiotic consumption data in vaccine studies, reaching consensus on design and implementation of specific studies aimed at characterising the role of specific vaccines against AMR or determining how to better use data from bacterial disease surveillance networks and routine monitoring of vaccine use. Also for further consideration is the systematic incorporation of AMR impact in vaccine-related decision-making in industry, financing bodies, regulatory and policy forums, including the creation of incentives for the development of future vaccines or better use of available vaccines with impact on AMR. Both WHO and Gavi have an important role in supporting coordinated action and rational decision-making in vaccine research and implementation.\n\n\nEthical approval and consent\n\nThis research is exempt from ethical approval. It involves the use of non-sensitive and anonymous surveys and interview procedures with participants that are not defined as “vulnerable” and participation will not induce undue psychological stress or anxiety. All participants gave a written consent to participate in the study and all data was anonymised prior to submission.\n\n\nData availability\n\nFigshare: AMR GAVI VIS underlying data. https://doi.org/10.6084/m9.figshare.972422933\n\nThis project contains the following underlying data:\n\nRaw data.xlsx (This is a raw data file with individual answers from study participants)\n\nFigshare: AMR GAVI VIS Extended data. https://doi.org/10.6084/m9.figshare.923211532\n\nThis project contains the following extended data:\n\nBackground info for authors.xlsx (The excel file that contains the background information that was sent to all participants in the study; the questionnaire that asked participants to enter the vaccine and criteria scores; and participants anonymised details)\n\nEmail body.pdf (The body of an email that was sent to all participants in the study)\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "Acknowledgements\n\nWe thank the experts who agreed to participate in this value attribution exercise. We also thank the following individuals for critical insights on the draft manuscript: Elizabeth Klemm, Lee Hampton, Duncan Graham-Rowe, Daniel Feikin.\n\nDisclaimer: The opinions expressed herein are those of the authors and do not necessarily reflect the views and decisions of the World Health Organization or Gavi, the Vaccine Alliance.\n\n\nReferences\n\nReview on Antimicrobial Resistance: Tackling Drug-resistant infections globally: Final Report and Recommendations. 2016. Reference Source\n\nUnited Nations: Draft political declaration of the high-level meeting of the General Assembly on antimicrobial resistance. 2016. Reference Source\n\nWorld Health Organisation: Global Action Plan on Antimicrobial Resistance. 2015. Reference Source\n\nTacconelli E, Carrara E, Savoldi A, et al.: Discovery, research, and development of new antibiotics: the WHO priority list of antibiotic-resistant bacteria and tuberculosis. Lancet Infect Dis. 2018; 18(3): 318–27. PubMed Abstract \| Publisher Full Text\n\nLaxminarayan R, Matsoso P, Pant S, et al.: Access to effective antimicrobials: A worldwide challenge. Lancet. 2016; 387(10014): 168–75. PubMed Abstract \| Publisher Full Text\n\nThe Review on Antimicrobial Resistance: Vaccines and Alternative Approaches: Reducing our Dependence on Antimicrobials. London, 2016. Reference Source\n\nClift C, Salisbury DM: Enhancing the role of vaccines in combatting antimicrobial resistance. Vaccine. 2017; 35(48 Pt B): 6591–3. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChatham House Centre on Global Health Security: The Value of Vaccines in the Avoidance of Antimicrobial Resistance, Meeting Summary. 2017. Reference Source\n\nKingwell K: Vaccines take a shot at antimicrobial resistance. Nat Rev Drug Discov. 2018; 17(4): 229–31. PubMed Abstract \| Publisher Full Text\n\nJansen KU, Knirsch C, Anderson AS: The role of vaccines in preventing bacterial antimicrobial resistance. Nat Med. 2018; 24(1): 10–9. PubMed Abstract \| Publisher Full Text\n\nKennedy DA, Read AF: Why the evolution of vaccine resistance is less of a concern than the evolution of drug resistance. Proc Natl Acad Sci U S A. 2018; 115(51): 12878–12886. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAtkins KE, Lafferty EI, Deeny SR, et al.: Use of mathematical modelling to assess the impact of vaccines on antibiotic resistance. Lancet Infect Dis. 2018; 18(6): e204–e213. PubMed Abstract \| Publisher Full Text\n\nKyaw MH, Lynfield R, Schaffner W, et al.: Effect of introduction of the pneumococcal conjugate vaccine on drug-resistant Streptococcus pneumoniae. N Engl J Med. 2006; 354(14): 1455–63. PubMed Abstract \| Publisher Full Text\n\nvon Gottberg A, de Gouveia L, Tempia S, et al.: Effects of vaccination on invasive pneumococcal disease in South Africa. N Engl J Med. 2014; 371(20): 1889–99. PubMed Abstract \| Publisher Full Text\n\nLaxminarayan R, Matsoso P, Pant S, et al.: Access to effective antimicrobials: a worldwide challenge. Lancet. 2016; 387(10014): 168–75. PubMed Abstract \| Publisher Full Text\n\nAdam HJ, Richardson SE, Jamieson FB, et al.: Changing epidemiology of invasive Haemophilus influenzae in Ontario, Canada: evidence for herd effects and strain replacement due to Hib vaccination. Vaccine. 2010; 28(24): 4073–8. PubMed Abstract \| Publisher Full Text\n\nKwong JC, Maaten S, Upshur RE, et al.: The effect of universal influenza immunization on antibiotic prescriptions: an ecological study. Clin Infect Dis. 2009; 49(5): 750–6. PubMed Abstract \| Publisher Full Text\n\nSevilla JP, Bloom DE, Cadarette D, et al.: Toward economic evaluation of the value of vaccines and other health technologies in addressing AMR. Proc Natl Acad Sci U S A. 2018; 115(51): 12911–12919. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBerkley S (Gavi): Report from Gavi, the Vaccine Alliance Meeting of the Strategic Advisory Group of Experts on Immunisation (SAGE). 2018. Reference Source\n\nGavi: Vaccine Investment Strategy 2018. (accessed July 26, 2019). Reference Source\n\nGlobal Burden of Disease Collaborative Network: Global Burden of Disease Study 2016 (GBD 2016) Population Estimates 1950-2016. Seattle, United States: 2017. Reference Source\n\nAtkins KE, Lafferty EI, Deeny SR, et al.: Use of mathematical modelling to assess the impact of vaccines on antibiotic resistance. Lancet Infect Dis. 2018; 18(6): e204–13. PubMed Abstract \| Publisher Full Text\n\nWorld Health Organization (WHO): Typhoid fever – Islamic Republic of Pakistan 2018. (accessed February 19, 2019). Reference Source\n\nAndrews JR, Baker S, Marks F, et al.: Typhoid conjugate vaccines: a new tool in the fight against antimicrobial resistance. Lancet Infect Dis. 2019; 19(1): e26–e30. PubMed Abstract \| Publisher Full Text\n\nWHO: Q&A on the malaria vaccine implementation programme (MVIP). 2019. Reference Source\n\nWorld Health Organization (WHO): Q&A on artemisinin resistance. 2018. Reference Source\n\nKiemde F, Spijker R, Mens PF, et al.: Aetiologies of non-malaria febrile episodes in children under 5 years in sub-Saharan Africa. Trop Med Int Health. 2016; 21(8): 943–55. PubMed Abstract \| Publisher Full Text\n\nGosling R, von Seidlein L: The Future of the RTS,S/AS01 Malaria Vaccine: An Alternative Development Plan. PLoS Med. 2016; 13(4): e1001994. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWellcome Trust, BCG: Vaccines to tackle drug resistant infections. 2018. Reference Source\n\nWorld Organisation for Animal Health: AHG on Prioritisation of diseases for which vaccines could reduce antimicrobial use in cattle, sheep, and goats. Paris, 2015. Reference Source\n\nMarsh K, Ijzerman M, Thokala P, et al.: Multiple Criteria Decision Analysis for Health Care Decision Making--Emerging Good Practices: Report 2 of the ISPOR MCDA Emerging Good Practices Task Force. Value Health. 2016; 19(2): 125–37. PubMed Abstract \| Publisher Full Text\n\nHasso-Agopsowicz M: AMR GAVI VIS Extended data. figshare. Dataset. 2019. http://www.doi.org/10.6084/m9.figshare.9232115\n\nHasso-Agopsowicz M: AMR GAVI VIS underlying data. figshare. Dataset. 2019. http://www.doi.org/10.6084/m9.figshare.9724229" }	[ { "id": "54296", "date": "01 Oct 2019", "name": "William Hausdorff", "expertise": [ "Reviewer Expertise Infectious disease epidemiology and vaccine development" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nBased on a survey of 18 global experts in antimicrobial resistance, this article recounts their perceptions of the value relative to AMR of investments in specific vaccines currently underway or being considered by GAVI. As such, it provides 2 types of information that are valuable: 1) expert views of the relative importance of pre-defined criteria by which to assess vaccine value vis a vis AMR and 2) expert assessments of specific vaccines.\nThe paper is nicely written, but there are a few places where greater clarity would help make interpretation of the results more straightforward.\n\nSpecifically, I remain unclear whether experts were assessing the vaccines based on their potential value or based on their current value. For example, current pneumococcal conjugate vaccines, while they all but eliminate antimicrobial non-susceptible invasive disease caused by 10-13 vaccine serotypes, don't prevent such disease caused by the other 80+ non-vaccine serotypes - and many of those can be antibiotic non-susceptible. Is the high valuation of PCVs based on the current PCVs or on a future common antigen vaccine/25+ valent conjugate? Similarly, the current RTS,S vaccine has real but limited efficacy - is the high valuation of malaria due to this vaccine or only to a theoretically 95% effective vaccine?\nSecondly, it was unclear to me whether the high priority given to a vaccine's ability to prevent mortality and morbidity was related to overall disease burden caused by that pathogen, or the magnitude of the antimicrobial non-susceptible disease burden caused by that pathogen, or the potential antimicrobial non-susceptible disease burden. This would help in interpreting the relatively high priority given to Hib-containing pentavalent vaccine, for example. And I'm not sure how much pneumococcal mortality is currently due to antimicrobial resistant pneumococci.\nIf the answer is that no further clarity is possible due to ambiguous wording of the questions, then such ambiguity could be pointed out as something to avoid in the future.\nAn additional limitation that should be highlighted is that all of the 18 experts queried appear to be from high income countries, with the exception of one (or more?) Indian expert.\nIt was unclear whether the outlier data points in Figures 1 and 2 were included in the calculations of the medians and means. It seems a bit strange that the error bars said to denote minimum and maximum values do not include those outliers; that means the error bars should be more accurately described as a CI covering X% of the data.\nMore minor comments to address to improve clarity:\nPage 3, Background, mentions \"a similar post-introduction study in South Africa\" of PCV, but the specific results described for that study do not provide any evidence of impact against AMR.\n\nPage 8, first few lines would seem to imply that 118,000 children mentioned all received typhoid polysaccharide vaccine, but in reality they received typhoid conjugate vaccine.\n\nAlso on page 8, it is unclear to me how the Chatham House prioritization of tuberculosis, typhoid, and influenza is \"broadly consistent\" with the results presented here.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "54549", "date": "21 Oct 2019", "name": "David M. Salisbury", "expertise": [ "Reviewer Expertise Vaccines", "vaccinology", "immunisation programme management", "polio eradication", "public health." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study takes a number of infections for which vaccine interventions exist or may exist, considers their relevance to AMR and takes into account the vaccines that have been prioritised for investment by GAVI. A number of experts have been asked to prioritise these vaccines against criteria that were predetermined and scored against AMR importance elements. This prioritisation exercise is very similar to those previously undertaken by and on behalf of the WHO Strategic Advisory Group of Experts on immunisation (SAGE). The results of these previous exercises, as this one, were driven essentially by mortality, and to a lesser degree, by morbidity: the results are not surprising irrespective of the complexity of the scoring system. Meanwhile, some confidence can be derived from this subjective process that the GAVI vaccine priorities remain relevant from an AMR perspective.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1685
https://f1000research.com/articles/8-52/v1	14 Jan 19	{ "type": "Research Article", "title": "Conserved structure of the NPR1 gene distal promoter isolated from a chili pepper (Capsicum annuum L.) in West Sumatera", "authors": [ "Jamsari Jamsari", "Maythesya Oktavioni", "Bastian Nova", "Ifan Aulia Candra", "Alfi Asben", "Lily Syukriani", "Maythesya Oktavioni", "Bastian Nova", "Ifan Aulia Candra", "Alfi Asben", "Lily Syukriani" ], "abstract": "Background: The non-expressor of pathogenesis related gene 1 (NPR1) protein is one of the key regulators in the systemic acquired resistance plant defence system. The cis-acting elements of its distal promoter gene are characterized by salicylic acid inducing elements such as the W-box, RAV1AAT and ASF1, accompanied with enhancer and silencer elements. This study was aimed to isolate and characterize the distal promoter sequence of the NPR1 gene (PD_CbNPR1) from the chili pepper (Capsicum annuum L.) genotype Berangkai, a local genotype known to produce large yields, but is susceptible to viral infection. Elucidating its sequence structure will open a broad range of possibilities to engineer the NPR1 gene expression which is important to improve chili pepper resistant. Methods: PCR-based cloning combined with a primer walking strategy was applied in this study. The BioEdit tool was used to edit the sequence and verify sequence integrity, while homology analysis was conducted with BLASTn searching. Identification of a cis-acting element was detected by PLACE. Results: Isolation of the complete distal promoter sequence of PD_CbNPR1 produced a fragment 5,950 bp in size. BLASTn search analysis indicated that PD_CbNPR1 sequence is highly conserved (99% homology) showing only a single nucleotide polymorphism (SNP) (base substitution) compared with its reference sequence. Analysis using PLACE tools successfully identified nine cis-acting elements containing a W-box, WLE1, RAV1AAT, TATA-box, CAAT-box, GARE and GT1 with multi repeats and diverse motives, as well as enhancer and silencer elements, which is characterized by a CCAAT-box and GAGAAATT pattern, respectively. Conclusion: The distal promoter of the NPR1 gene is highly conserved, showing only one SNP caused by one base substitution event.", "keywords": [ "CCAAT-box", "distal promoter", "NPR1 gene", "RAV1AAT", "W-box" ], "content": "Introduction\n\nThe non-expressor of pathogenesis related gene 1 (NPR1) protein is a main regulator in the systemic acquired resistance response of many plants. Over expression by modifying distal promoter in the W-box element of the OsNPR1 gene in rice could increase its resistance against Xanthomonas oryzae pv. Oryzae up to 4.3 times (Hwang & Hwang, 2010). Similar results were reported by Zhong et al. (2015), who modified the RAV1 element on the GhNPR1 which increased Gladiolus hybridus resistance against Curvularia gladioli by up to 18.6 times compared to control. Based on those studies, we expect more prospects of the NPR1 gene promoter in the improvement of plant resistance against many pathogens.\n\nHere we report characteristics of the distal promoter segment of the NPR1 gene isolated from chili pepper (Capsicum annuum) genotype Berangkai, a local genotype potentially produces more yields compared to other genotypes cultivated in West Sumatera.\n\n\nMethods\n\nHealthy young leaves collected from chili pepper genotype Berangkai was used as the plant material in this study. The chili pepper genotype Berangkai was grown in a greenhouse and maintained for 8 weeks before being used for DNA isolation. Genomic DNA isolation was performed using protocol as described by Jamsari & Pedri (2013).\n\nPCR-based cloning combined with a primer walking approach was applied for the isolation complete segment of distal promoter of the NPR1 gene. All PCR reactions in this study were performed using the KOD-Plus-Neo kit provided by Toyobo-Japan. Chromosome 7 of Capsicum annuum cv. Zunla-1 (Qin et al., 2014), accession number: NC_029983.1, was used as the reference sequence. Primer combinations were designed to cover the interval segment spanning from base 112.600.897 to 112.606.847 on the reference sequence (Table 1). All primers used in this study are listed in Table 1. First-step isolation was started from both termini and continued with successive isolation until both termini formed a complete contig of the target segment. The initial PCR condition was started with 95°C for 3 minutes for pre-denaturation and amplified in two different loops using 14 and 24 cycles. The first 14 cycles were performed using a touch down steps started by denaturation at 95°C for 30 seconds. Annealing was initiated with 70°C for 30 seconds and gradually decreased 1°C for every cycle before finally elongated in 72°C for 2 minutes. The next 24 cycles started with denaturation at 95°C for 30 seconds with annealing at 55°C for 30 seconds and elongation of at 70°C for 2 minutes and elongated for 72°C for 2 minutes. Final extension was maintained at 72°C for 5 minutes. PCR reaction was performed in 50 µl of final volume. PCR composition was set according to manufacturer’s recommendation, containing 5 µl 10x Buffer KOD-Plus-Neo, 5 µl of 2 mM dNTPs, 3 µl of 25 mM MgSO4, 1.5 µl of 10 ng/µl of each primer, 1 µl of KOD-Plus-Neo (Toyobo, Japan) and 1 µl of 10 ng/µl of DNA template. The volume was filled with 32 µl of nuclease-free water.\n\nThe PCR product generated from each walking step was processed to sequencing reaction using its both forward and reverse primers. All sequencing processes were performed by 1st BASE-Singapore. A number of bioinformatic tools were applied during analysis of the sequence. Trimming, editing and building of the sequence contig were performed using BioEdit v7.2.5 (Hall, 1999). Homology search of the sequence with all available sequences deposited in the NCBI database was run using the BLASTn tool (Altschul et al., 1990). The cis-acting elements were identified using PLACE, developed by Higo et al. (1999).\n\n\nResults and discussion\n\nWe successfully isolated the distal promoter of the NPR1 gene from our local chili pepper genotype Berangkai, exhibiting 5,950 bp in size and designated PD_CbNPR1. The fragment was isolated from two steps round of walking via PCR-based cloning. In the first step, the whole distal promoter region was expected to be covered using two primer combinations (F1 F/R and F2 F/R). However, after verification of the sequence data using BioEdit, the upstream primer combination (F1 F/R) successfully produced a contig spanning only 1,963 bp, while the downstream primer combination (F2 F/R) was skipped for the sequencing process. In the second round of primer walking, four new primer combinations (F1.1-F/R, F1.2-F/R, F2.1-F/R, and F2.2-F/R) (Table 1) were designed in order to extend sequence coverage. Combining all verified sequences obtained from the first and the second round of primer walking successfully produced a contig with a size of 5,950 bp (Figure 1).\n\nBase position of start and end segment are shown by number on each terminus. Length of verified sequence of each segment is shown by number in bracket on each segment.\n\nHomology analysis between PD_CbNPR1 and its reference sequence via BLAST search showed 99% homology. The data was verified by only one single nucleotide polymorphism (SNP) shown as a substitution event at the position -6,335 from the ATG start codon (Figure 2). This finding clearly indicated that both distal promoters are highly conserved, even they originated from distantly related regions.\n\nPLACE analysis successfully indicated 9 cis-acting elements with multi repetition (Figure 3). All 9 cis-acting elements contained the W-box (15), WLE1 (W-box like elements) (8), RAV1AAT (20), TATA-box (22), CAAT-box (26), GARE (1), GT1 (20), Enhancer (4) and Silencer (1).\n\nColours indicate each element as labelled below the figure.\n\nThe W-Box element binds the WRKY protein, acting as a transcription factor during expression of NPR1 (Yu et al., 2010). Mutation in this element in Arabidopsis thaliana delayed the NPR1 expression induced by salicylic acid (SA). PLACE analysis indicated that PD-CbNPR1 is characterized by three consensus sequences: TTGAC, TGACY, and TGACT. The TGACY pattern could also be found in the promoter of ERF3 isolated from tobacco (Nishiuchi et al., 2004), while TGACT could be found in the promoter of the Iso1 barley gene (Sun et al., 2003). The WLE1 (W-box like elements) with TGACA pattern has an analog function with the W-box identified in the promoter of OsNPR1 (Hwang & Hwang, 2010).\n\nAnother cis-acting element found in PD_CbNPR1 is RAV1AAT, which acts as binding site for protein RAV1 (Hwang & Hwang, 2010). The RAV1 protein is a transcription factor associated with some pathogenesis-related genes (Sohn et al., 2006). Two consensus patterns of RAV1AAT elements found in PD_CbNPR1 are CAACA and TGTTG. The CAACA pattern was also found by Kagaya et al. (1999) in Arabidopsis thaliana, while the TGTTG pattern is similar to the OsNPR1 promoter (Hwang dan Hwang, 2010). Notably, the ASF1 element is absent in the PD_CbNPR1. The ASF1 element has been previously reported by Hwang & Hwang (2010) and Zhong et al. (2015) as a cis-acting for common promoter associated with pathogen infection and SA induction.\n\nPLACE analysis identified 22 putative TATA-boxes in our PD_CbNPR1, which could be classified into TATA box 3, TATA box 4, TATA box 5 and TATA box OspaI. During transcription, the TATA-box binds with RNA polymerase II and many transcription factors like TFIID, TFIIA, TFIIF, TFIIE, and TFIIH to build the TATA-box complex.\n\nAnother interesting element found in PD_CbNPR1 is the CAAT-box (26 repeats) which is associated with the regulation of many genes involved in pathogen infection (Imran et al., 2016). The CAAT-box consensus sequence found in PD_CbNPR1 is characterized with CAAT which is similar that have been found in promoter of the legA gene from Pisum sativum (Shirsat et al., 1989).\n\nWe also found a gibberelline-responsive element (GARE) motif, which has been previously reported by Ogawa et al. (2003) as the binding site with some transcription factors induced by gibberellic acid. The consensus sequence of thoses motifs is TAACAAR. Another cis-acting element induced by abiotic factors such as salt, light and pathogenesis is GT1 motif (Biłas et al., 2016; Park et al., 2004). The identified GT1 element in PD_CbNPR1 is characterized by a consensus GRWAAW motif and present in 20 repeats. They could be classified into two forms, GT1CONSENSUS and GT1GMSCAM4, as described previously by Terzaghi & Cashmore (1995). The last GT1 class, GT1GMSCAM4, is initially identified by Park et al. (2004) with consensus pattern GAAAAA.\n\nPLACE analysis also successfully identified four CCAAT-box motifs, which is commonly reported as an enhancer of transcription (Thonpho et al., 2013). This element is able to up-regulate the transcription and replication rate of eukaryotic genes by binding many types of transcription factors (Biłas et al., 2016). da Silva et al. (2013) previously reported that a greater number of CCAAT-boxes could improve promoter capability in increasing of replication rate by binding efficiently with transcription factors. In contrast with the enhancer, we found only one motif of GAGAAATT which is also known as silencer (Lai et al. (2009). This element is reported to be able to down-regulate the kinesin-like protein 1 (AtKP1) gene in Arabidopsis thaliana by up to 80.9%.\n\n\nConclusion\n\nWe successfully isolated the distal promoter of the NPR1 gene from chili pepper genotype Berangkai (designated PD_CbNPR1) spanning 5,950 bp. The PD_CbNPR1 contain 9 cis-acting elements: W-box, WLE1, RAV1AAT, TATA-box, CAAT-box, GARE, GT1, enhancer and silencer elements. Engineering of the cis-acting elements may have future prospects, particularly in improving chili pepper resistance against biotic and abiotic stresses by up- and down-regulating NPR1 gene expression.\n\n\nData availability\n\nThe sequence of the distal promoter of the NPR1 gene sequenced in this study has been deposited in GenBank under accession number MK281381; http://identifiers.org/ncbigi/GI:1547604514.", "appendix": "Grant information\n\nThis study was funded by Universitas Andalas–Padang, West Sumatera, Indonesia, via Professorship Cluster Research Grants, Fiscal year 2018, Contract No.: 18/UN.16.17/PP.RGB/LPPM 2018.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nAltschul SF, Gish W, Miller W, et al.: Basic local alignment search tool. J Mol Biol. 1990; 215(3): 403–410. PubMed Abstract \| Publisher Full Text\n\nBiłas R, Szafran K, Hnatuszko-Konka K, et al.: Cis-regulatory elements used to control gene expression in plants. Plant Cell Tiss Organ Cult. 2016; 127(2): 269–287. Publisher Full Text\n\nda Silva JS, Slowik R, Bicalho Mda G: Considerations on regulatory sequences of the distal promoter region of the HLA-G gene. Hum Immunol. 2013; 74(4): 473–477. PubMed Abstract \| Publisher Full Text\n\nHall TA: BioEdit a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucl Acids Symp Ser. 1999; 41: 95–98. Reference Source\n\nHigo K, Ugawa Y, Iwamoto M, et al.: Plant cis-acting regulatory DNA elements (PLACE) database: 1999. Nucleic Acids Res. 1999; 27(1): 297–300. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHwang SH, Hwang DJ: Isolation and characterization of the rice NPR1 promoter. Plant Biotechnol Rep. 2010; 4(1): 29–35. Publisher Full Text\n\nImran QM, Falak N, Hussain A, et al.: Nitric Oxide Responsive Heavy Metal-Associated Gene AtHMAD1 Contributes to Development and Disease Resistance in Arabidopsis thaliana. Front Plant Sci. 2016; 7: 1712. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJamsari J, Pedri J: Complete nucleotide sequence of DNA A-like genome and DNA-β of monopartite pepper yellow leaf curl virus, a dominant Begomovirus infecting Capsicum annuumin West Sumatera Indonesia. Asian J Plant Pathol. 2013; 7(1): 1–14. Publisher Full Text\n\nKagaya Y, Ohmiya K, Hattori T: RAV1, a novel DNA-binding protein, binds to bipartite recognition sequence through two distinct DNA-binding domains uniquely found in higher plants. Nucleic Acids Res. 1999; 27(2): 470–478. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLai C, Xiong J, Li X, et al.: A 43-bp A/T-rich element upstream of the kinesin gene AtKP1 promoter functions as a silencer in Arabidopsis. Plant Cell Rep. 2009; 28(5): 851–860. PubMed Abstract \| Publisher Full Text\n\nNishiuchi T, Shinshi H, Suzuki K: Rapid and transient activation of transcription of the ERF3 gene by wounding in tobacco leaves: possible involvement of NtWRKYs and autorepression. J Biol Chem. 2004; 279(53): 55355–55361. PubMed Abstract \| Publisher Full Text\n\nOgawa M, Hanada A, Yamauchi Y, et al.: Gibberellin biosynthesis and response during Arabidopsis seed germination. Plant Cell. 2003; 15(7): 1591–1604. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPark HC, Kim ML, Kang YH, et al.: Pathogen- and NaCl-induced expression of the SCaM-4 promoter is mediated in part by a GT-1 box that interacts with a GT-1-like transcription factor. Plant Physiol. 2004; 135(4): 2150–2161. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nQin C, Yu C, Shen Y, et al.: Whole-genome sequencing of cultivated and wild peppers provides insights into Capsicum domestication and specialization. Proc Natl Acad Sci USA. 2014; 111(14): 5135–5140. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShirsat A, Wilford N, Croy R, et al.: Sequences responsible for the tissue specific promoter activity of a pea legumin gene in tobacco. Mol Gen Genet. 1989; 215(2): 326–331. PubMed Abstract \| Publisher Full Text\n\nSohn KH, Lee SC, Jung HW, et al.: Expression and functional roles of the pepper pathogen-induced transcription factor RAV1 in bacterial disease resistance, and drought and salt stress tolerance. Plant Mol Biol. 2006; 61(6): 897–915. PubMed Abstract \| Publisher Full Text\n\nSun C, Palmqvist S, Olsson H, et al.: A novel WRKY transcription factor, SUSIBA2, participates in sugar signaling in barley by binding to the sugar-responsive elements of the iso1 promoter. Plant Cell. 2003; 15(9): 2076–2092. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTerzaghi WB, Cashmore AR: Light-regulated transcription. Plant Mol Biol. 1995; 46(1): 445–474. Publisher Full Text\n\nThonpho A, Rojvirat P, Jitrapakdee S, et al.: Characterization of the distal promoter of the human pyruvate carboxylase gene in pancreatic beta cells. PLoS One. 2013; 8(1): e55139. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYu J, Wang XY, Wei Q, et al.: Identification of regulatory cis-elements upstream of AtNPR1 that are responsive to probenazole treatment in transgenic tobacco plants. J Plant Biol. 2010; 53(4): 282–290. Publisher Full Text\n\nZhong X, Xi L, Lian Q, et al.: The NPR1 homolog GhNPR1 plays an important role in the defense response of Gladiolus hybridus. Plant Cell Rep. 2015; 34(6): 1063–1074. PubMed Abstract \| Publisher Full Text" }	[ { "id": "43017", "date": "28 Jan 2019", "name": "Minghui Lu", "expertise": [ "Reviewer Expertise plant tolerance to abiotic stresses especially high temperature" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn the manuscript “Conserved structure of the NPR1 gene distal promoter isolated from a chili pepper (Capsicum annuum L.) in West Sumatera”, Dr. Jamsari Jamsari and co-authors offer us the structure of the distal promoter of the NPR1 gene, which will lay a foundation for the understanding of the mechanisms of pepper resistance to viral infection. But there are some issues that need to be explained:\nThe PD_CbNPR1 is cloned by gene walking approach, which needs to be confirmed by PCR amplification with a full-length primer pair.\n\nThe authors suggest that the structure of PD_CbNPR1 is conservative based on the comparison with the reference gene (Figure 2) and on the cis-element distribution (Figure 3). I think that this is not enough. The comparison of nucleotide sequences among distal promoters from several different plant species should be performed.\n\nThe PD_CbNPR1 is amplified from a pepper variety with viral susceptibility. I suggest the authors compare the PD sequences between virus susceptible and resistant varieties, which is beneficial to understanding the mechanisms of pepper resistance to viral infection.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4919", "date": "24 Sep 2019", "name": "Jamsari Jamsari", "role": "Author Response", "response": "Reviewer's Comment (RC): The PD_CbNPR1 is cloned by gene walking approach, which needs to be confirmed by PCR amplification with a full-length primer pair.Author's Response (AR): Thank you for your comments. The gene walking approach we applied was based on the reason, that the length of the putative distal promoter NPR1 gene region (5,950 bp) could not be sequenced in one step, due to sequencer machine capacity, which is normally in average about 500 bp. That's why we have chosen this strategy and considered it to be the most appropriate to obtain the full length of the putative sequence. Producing a PCR fragment of 5,950 bp in size is indeed possible by using a specific DNA Polymerase i.e.: the KOD-Plus-Neo that could amplify up to 24 kb (according to the manufacturer’s claim), but the full-length fragment (5,950 bp) still cannot be sequenced in one step read due to limited reading capacity of the sequencing machine. To confirm the validity of every single nucleotide data, we confirmed by at least two overlapping validated segment each other. This argument will be added in the 2nd version.=======================================================================RC: Comment 2: The comparison of nucleotide sequences among distal promoters from several different plant species should be performed.AR: We tried already to BLAST our nucleotide sequence in order to get more sequence data with significant identity. Unfortunately the result exhibited no significant homology with another promotor sequence available in the NCBI database, indicating a limitation of promoter sequence availability in the public database. The only promoter sequence which showed homology is the promoter region of Capsicum annuum pathogenesis related protein-1 (PR-1) gene (DQ201633.1) published by Lee et al., (2006). However, the comparable nucleotide of both sequences spanned only 180 bp. We put this discussion in the 2nd version.=======================================================================RC: Compare the PD sequences between virus susceptible and resistant varieties, which is beneficial to understanding the mechanisms of pepper resistance to viral infection.AR: Indeed your suggestion is true and we agree 100%. Unfortunately the resistant genotype, is not available in our collection so far, so comparison of both two genotypes is not possible to be performed.We put an additional sentence regarding this issue in the first paragraph of the methods." } ] }, { "id": "47197", "date": "02 May 2019", "name": "Santy Peraza-Echeverria", "expertise": [ "Reviewer Expertise Molecular Plant-Pathogen Interaction and Plant Biotechnology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study reports the cloning and bioinformatic analysis of the putative promoter region of PD_CbNPR1 from chili pepper. This putative promoter region was isolated from a local chili pepper genotype (Berangkai) known to produce large yields but susceptible to viral infections. The putative promoter sequence of PD_CbNPR1 could be useful for enhancing the immune system of the Berangkai genotype.\n\nI have the following observations and comments:\nIndicate the Genbank accession number of the NPR1 sequence used to BLAST the chili pepper genome. Also, perform a phylogenetic analysis to determine whether the chili NPR1 sequence used to isolate PD_CbNPR1 belongs to the NPR1 clade (e.g., Arabidopsis has five NPR1 homologs: NPR2, NPR3, NPR4, NPR5 and NPR6. NPR1 and NPR2 form a single clade).\n\nCheck the concept of homology. It is wrong to say \"99% homology\".\n\nThe KOD-Plus-Neo kit is designed for long range PCR (up to 24 kb from human genomic DNA), so it is not clear why the amplification of the full-length putative promoter region (5950 bp single amplicon) was not possible.\n\nThe search for putative cis-acting regulatory elements in the putative promoter of PD_CbNPR1 should also include other databases such as:\n\na. PlantCare (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/)\n\nb. PlantPAN 3.0 (http://plantpan.itps.ncku.edu.tw)\n\nFigure 1: the length of the core promoter should be indicated along with the 5’UTR.\n\nThe transcription start site should be provided as part of the characterization of the promoter region.\n\nA transient expression analysis using GUS as reporter gene would be useful to determine whether this putative promoter is functional.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4920", "date": "24 Sep 2019", "name": "Jamsari Jamsari", "role": "Author Response", "response": "Reviewer's Comment (RC): Indicate the Genbank accession number of the NPR1 sequence used to BLAST the chili pepper genome. Also, perform a phylogenetic analysis to determine whether the chili NPR1sequence used to isolate PD_CbNPR1 belongs to the NPR1 clade (e.g., Arabidopsis has five NPR1 homologs: NPR2, NPR3, NPR4, NPR5 and NPR6. NPR1 and NPR2 form a single clade).Author’s Response (AR): We analyzed already with some additional cDNA sequences of the NPR1 and found interesting new insight. However, our NPR1 cDNA sequence showed no significant similarity with the NPR1 cDNA sequence isolated from Arabidopsis, including its family (homologs) i.e.: NPR2, NPR3, NPR4, NPR5 and NPR6. Thus, clustering analysis separated our NPR1 from Arabidopsis clade. Interestingly, we found 43 significant identity with our NPR1 cDNA sequence from many different solanaceae, i.e.: Capsicum annuum (NM_001325099.1), Capsicum chinense (AM900559.1), Solanum lycopersicum (KX198701.1, NM_001247629.2), Nicotiana sp. (DQ837218.1, AF480488.1) Carica papaya (XM_022041103.1, AY550242.1) and some others. Tree analysis showed that our NPR1-Berangkai cDNA sequence clustered to similar clade with AM900559.1 and NM_001325099.1 and other three solanaceae (S. lycopersicum-KX198701.1, NM_001247629.2; S. tuberosum-XM_006357647.2; S. pennellii-XM_015227358.2 and S. torvum-KJ995663.1). We put this data in the 2nd version as suggested.========================================================RC: Check the concept of homology. It is wrong to say \"99% homology\".AR: Thank you for the correction, and we already changed this concept in the 2nd version.========================================================RC: The KOD-Plus-Neo kit is designed for long range PCR (up to 24 kb from human genomic DNA), so it is not clear why the amplification of the full-length putative promoter region (5950 bp single amplicon) was not possible.AR: You are right, and this is also our consideration in the beginning of the project. However, due to the sequencer read capacity used by the company handling for sequencing process which is limited for about only 500 bp on average, one step read is not possible to be applied. Even though the KOD-Plus-Neo could amplify up to 24 kb according to the manufacturer’s claim, the full-length fragment (5,950 bp) still cannot be sequenced in one step read due to limited reading capacity of the sequencing machine. Confirmation of data validity from every single nucleotide was confirmed by at least two overlapping validated segments. We put this argument in the results and discussion.========================================================RC: The search for putative cis-acting regulatory elements in the putative promoter of PD_CbNPR1should also include other databases such as: a. PlantCare (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) b. PlantPAN 3.0 (http://plantpan.itps.ncku.edu.tw)AR: Thank you for your constructive suggestions. We did additional analysis using PlantCARE and PlantPAN, and found two additional cis-acting motifs: 1 TCA motif and 3 CGTA motis which could not be shown by PLACE. We put already these additional motifs in Figure 3 and included in the 2nd version.========================================================RC: Figure 1: the length of the core promoter should be indicated along with the 5’UTR.AR: We made a modification indicating length and the position of the core promoter and included this modification in Figure 1.========================================================RC: The transcription start site should be provided as part of the characterization of the promoter region.AR: We already put the position of the transcription start site (TSS) located up stream of the NPR1 gene position. Thanks a lot for the suggestion.========================================================RC: Transient expression analysis using GUS as reporter gene would be useful to determine whether this putative promoter is functional.AR: That’s really a good idea, however we are performing the functionality with a different approach. Instead of GUS reporter platform, we are using it in bacterial system. Currently we are working to use some of the cis-acting elements found in this research as a promoter for expression of the C1 (Replicase) gene isolated from geminivirus in a bacterial system. However, we need more time until we get the validated data as suggested. Thanks again for your critical and inspiring comments and suggestions." } ] } ]	1	https://f1000research.com/articles/8-52
https://f1000research.com/articles/8-1684/v1	24 Sep 19	{ "type": "Research Article", "title": "Stress urinary incontinence in relation to pelvic floor muscle strength and associated factors in the third trimester of pregnancy: A cross-sectional study", "authors": [ "Astrid Yunita", "Tyas Priyatini", "Tyas Priyatini" ], "abstract": "Background: Many predictors of stress urinary incontinence (SUI) during pregnancy have been investigated. However, no studies have specifically identified a cutoff for pelvic floor muscle (PFM) strength and associated factors that could predict SUI during pregnancy. The aim of this study was to identify the cutoff between PFM strength and SUI, late in the third trimester of pregnancy and associated factors in Indonesian women. Methods: A cross-sectional study was conducted involving 142 women with a pregnancy of 36–40 weeks of gestational age at the Obstetrics and Gynecology clinic of Tebet Subdistrict Hospital, Jakarta, Indonesia. The data were collected through a medical interview, Questionnaire for Urinary Incontinence Diagnosis, physical examination, perineometer, and cough test. Results: SUI was identified in 54.2% of the 142 women. PFM strength 25.5 cmH2O and estimated fetal weight (EFW) ³3,100 g were the factors with the greatest influence on SUI (odds ratio (OR) = 2.52, p = 0.021, and OR = 3.34, p = 0.001, respectively). For women with PFM strength £25.5 cmH2O and EFW ³3,100 g, the prediction for SUI was ~75.39%. Conclusion: Weakening of the PFMs and EFW influence SUI. The cutoff values identified for both variables may be helpful for predicting SUI late in pregnancy.", "keywords": [ "QUID", "Stress urinary incontinence", "pelvic floor muscle strength", "perineometer" ], "content": "Introduction\n\nStress urinary incontinence (SUI) is defined by the International Continence Society as the involuntary loss or leakage of urine1, and is the most common type of urinary incontinence (UI) in pregnant women. The severity of SUI tends to become more pronounced as the pregnancy progresses because of the growth of the uterus, hormonal influences, and biomechanical alterations of the pelvis. As a result, both muscle tone and muscle strength in the pelvic region may decrease during pregnancy2. The prevalence of SUI among pregnant women is variable (range 18.6% to 75%), increases with gestational age, and is typically worst in the third trimester, followed by the second and first trimesters3–5. SUI represents a major health problem that affects about 54.3% of all pregnant women and has a negative impact on quality of life and health-care expenditures6,7.\n\nPregnancy may be associated with a reduction in the strength of the pelvic floor muscles (PFMs), which may lead to reduced strength and supportive sphincter function, and eventually, to SUI. However, the exact mechanism responsible for pregnancy-related SUI is not well understood. By actively supporting the pelvic organs and closing the urethral sphincter by contracting6, the PFMs play an integral role in the maintenance of the continence mechanism. Studies of pregnant women with SUI have found significantly decreased PFM strength in incontinent compared with continent pregnant women8. SUI during pregnancy can be prevented and PFM strength can be improved by pelvic floor muscle exercises (PFMEs). Oliveira et al.4 reported that performing PFMEs significantly increased PFM pressure and strength during pregnancy, and another study reported that this effect was maintained postpartum9.\n\nIdentification of predictors of SUI is important for the development of preventive strategies. Many predictors of SUI during pregnancy have been investigated. Some studies have focused on obstetric and neonatal factors, whereas others have noted the importance of clinical factors5. However, no studies have focused on identifying a cutoff for PFM and associated obstetric factors related to SUI in pregnancy. The aims of this study were: (1) to investigate the relationships between SUI and obstetric factors during pregnancy and patient characteristics in Indonesian women, (2) to identify the clinical predictors of SUI, particularly PFM strength, in the third trimester of pregnancy using a digital examination perineometer, and (3) to develop a predictive model of SUI with cutoff points for use in practice for the early detection, and possible prevention, of modifiable risk factors that contribute to the development of SUI during pregnancy.\n\n\nMethods\n\nThis cross-sectional study was conducted in the Obstetrics and Gynaecology outpatient clinic of Tebet Sub District Hospital, Jakarta. The project was approved by the Medical Ethics Board of the Faculty of Medicine Universitas Indonesia (0317/UN2.F1/Ethics/2018).\n\nThe participants were enrolled between March and May 2018. They were recruited to the study during routine checkups where they were made aware of the study and asked to participate. Initially, 149 pregnant women in their third trimester (36–40 weeks) of pregnancy were enrolled in this study. Inclusion criteria included age between 20 and 45 years old and women in their third trimester of pregnancy (36–40 weeks gestational age). None of the women exhibited the exclusion criteria of urinary tract infection, previous surgery for pelvic dysfunction, previous UI, pregnancy complications (preeclampsia, eclampsia, lung edema, or sepsis), neurological dysfunction (mental retardation, dementia, Parkinson disease, multiple system atrophy, multiple sclerosis, myelodysplastic syndrome, spinal stenosis, disc diseases, previous trauma, or surgery of spinal cord). However, seven women were excluded because of the presence of diabetes, as shown in laboratory tests, resulting in a total of 142 participants. The goals of the study and the procedures indicated were explained to all participants, and those who agreed to participate signed a written informed consent form.\n\nDemographic characteristics such as age, gestational age, educational level, body mass index (BMI), weight, and height were recorded. Obstetric and neonatal characteristics and clinical factors were investigated as potential predictors of SUI. The obstetric and neonatal characteristics were parity, mode of delivery, and estimated fetal weight during pregnancy.\n\nSUI was identified in a medical interview that included the symptom-based Questionnaire for Urinary Incontinence Diagnosis (QUID)10, In this study, QUID were modified and validated into the Indonesian language (Cronbach’s alpha values = 0.965 for SUI and 0.962 for urge UI). The diagnosis of SUI was based on the symptoms of involuntary loss of urine on effort or physical exertion, or on sneezing or coughing, in accordance with the definition of the International Continence Society.\n\nFor the physical examination, the participant’s leg was positioned in lithotomy position at the examination table with a full bladder. The patient was then asked to cough, if there was a fluid loss from the urethra together with the cough, it was considered as a positive cough test. Furthermore, if no leakage was observed, the cough stress test was repeated in standing position with full bladder11,12.\n\nAfter taking these measurements, the obstetrician–gynecologist used a PeritronTM perineometer and digital palpation to measure the squeeze pressure as an indicator of the participant’s PFM strength during a maximum voluntary contraction. Pressure was measured in cmH2O13–15. The sequence for muscle testing involved three repetitions of maximum voluntary contractions lasting 3s each, with a 10-s rest period between each. This sequence was followed by a 2-min rest break. The highest of three voluntary PFM contractions was recorded and used in the analysis. Before performing these contractions, the participants were taught the correct way to perform the PFM contractions by focusing on the PFM without excessive coactivation of the gluteal, hip adductor, and rectus abdominal muscles, as verified by visual inspection and digital palpation. All clinical measurements were taken by an obstetrics and gynecology doctor, and the questionnaire was administered by a trained interviewer.\n\nThe data were recorded in case registration forms and were analyzed using the Statistical Package for the Social Sciences (version 17.0; SPSS Inc., Chicago, IL). The participants’ demographic characteristics, outcome variable, and scores on the QUID and cough test are presented descriptively. The chi-squared test for categorical data and the Mann–Whitney U test for interval data were used to compare participants with and without SUI in the analysis to identify potential predictors. Predictors with p < 0.25 were entered into the classification and regression analysis, a multivariate nonparametric procedure. The accuracy of the classifications of the predictive model was determined according to the sensitivity, specificity, and area under the receiver-operating characteristic (AUC-ROC) curve. A level of significance of 0.05 was set to test whether the AUC-ROC curve in the predictive model differed from 0.5.\n\n\nResults\n\nDuring the study period, a total of 149 gravidas were identified; all had attended the outpatient clinic, but seven were excluded because of the presence of diabetes as indicated by laboratory tests. The remaining 142 women were enrolled in the study; 66 were multiparous, 56 were primiparous, and 20 were nulliparous. The women were classified according to their prior deliveries as follows: vaginal delivery only (65.5% of participants), caesarean delivery only (11.3%), caesarean and vaginal deliveries (3.5%), vacuum extraction only (0.7%), vacuum extraction and vaginal deliveries (1.4%), and no prior delivery (16.2%). These modes of delivery showed that more than half of the participants (67.6%) had history of vaginal delivery (spontaneous and assisted vaginal delivery) and most of them was diagnosed with SUI 47.8%. Moreover, multiparous also play a pivotal role in SUI 60.6% compared to Primiparous and nulliparous (48.7%).\n\nThe prevalence of SUI was 54.2% based on QUID and positive cough test. QUID was designed to classify the type of urinary incontinence (SUI, urge urinary incontinence and mixed urinary incontinence), while the cough test strengthened the identification of women with SUI. Both measures ensured consistency and objectivity of approach in assessment, as explained before in data collection section. However, QUID without the presence of a positive cough test was still taken as SUI in this study (this research did not discuss the grade of severity of SUI).\n\nThe median age was 32 years (range 20–45 years), and the median gestational age was 38 weeks (range 36–40 weeks). Over half of women (46.5%) were included in the overweight criteria according to the Asia Pacific WHO classification and followed by 24.6% of women who were obese. The patients had a median PFM strength of 24.5 (14–90) cmH2O. Multiparous women had a higher prevalence of SUI (60,6%), compared to primiparous and nulliparous women collectively (48.7%).\n\nThe results of the bivariate analysis of the two groups (SUI and non-SUI) are presented in Table 1. Age, gestational age, BMI, constipation, and parity did not differ between groups. The estimated fetal weight (EFW) (p < 0.05) and PFM strength (p < 0.05) differed significantly between the groups. Gestational age, EFW, and parity were higher in the SUI than in the non-SUI group, and significant data showed that PFM strength was significantly lower in the SUI group.\n\n*Significant at 5% level. SUI, stress urinary incontinence.\n\nEstimated fetal weight, parity, and PFM were included in the multivariate analysis because they had p values < 0.25 in the first analysis. The results from adjusted multivariate logistic regression using the backward process are presented in Table 2. PFM strength and EFW were the most influential factors on SUI.\n\nOR, odds ratio; CI, confidence interval; PFM, pelvic floor muscle.\n\nThe AUC-ROC curve for PFM strength is shown in Figure 1A, and that for estimated fetal weight in Figure 1B. The best cutoffs for predicting the risk of SUI in pregnancy were a PFM strength of ≤25.5 cmH2O (sensitivity 60% and specificity 65%) and estimated fetal weight of 3,100 g (sensitivity 64.9 and specificity 66.2%). The results of the probability test of the contribution of EFW and PFM strength to SUI showed that the risk of SUI was 75.39% if the EFW was ≥3,100 g and PFM strength was ≤25.5 cmH2O. This was followed by values of 54.9% if the EFW was ≥3,100 g and PFM strength was >25.5 cmH2O, and 48% if the EFW was <3,100 g and PFM was ≤25.5 cmH2O.\n\n(A) ROC curve for PFM; area under the curve = 65.2%, sensitivity = 60%, and specificity = 65%. (B) ROC curve for estimated fetal weight; area under the curve = 65.2%, sensitivity = 64.9%, and specificity = 66%.\n\nUnderlying data are available from Harvard Dataverse16.\n\n\nDiscussion\n\nAlthough much has been written about childbirth as a predisposing risk factor to pelvic floor damage and SUI, well-conducted studies have pointed to the impact of pregnancy itself. This study demonstrated a high prevalence of SUI symptoms (54.2%) in the late third trimester. This finding is similar to those of Whitford et al.17, who reported a prevalence rate of SUI 54.3%, and Martins et al.18, who found an incidence range from 46.1% to 54%.\n\nIn the present study, the characteristic data explained that there were no difference between the age of the women in the SUI and non-SUI group; however, those aged >30 years was the median age in this study. Age is one of influencing factors for SUI; the study by Sangsawang5 found that pregnant women aged >30 years or who had their first pregnancy in their 30s were at high risk for SUI. Another study noted that women aged ≥30 years were susceptible to developing persistent and long-term SUI because of PFM defects19. The nerve function and the total number of striated muscle fibers in the urethral sphincter decrease with age at a rate of 2% a year and these changes can lead to a decrease in urethral closure pressure20–22. Despite the fact that the results of this study showed no age differences between the two groups, it could not conclude that age had no influence on SUI, and this needs to be further studied.\n\nIn the analysis of parity, nulliparous and primiparous women had the lowest prevalence of UI (51.3%), and the prevalence of SUI increased with parity to 60.6% in multiparous women. Spontaneous vaginal delivery was more strongly associated with SUI than delivery by other routes. These findings are similar to those of other authors, which have shown that age at childbirth, gestational age, parity2,23, and mode of previous births24,25 influence the occurrence of SUI during pregnancy. However, the bivariate analysis of our data showed that age, gestational age, parity, BMI, and constipation had no significant relationship with SUI (p > 0.05). Further studies are needed with a larger number of participants.\n\nPregnancy- and delivery-related factors are considered to be the main risk factors for SUI development during pregnancy because they can increase intra-abdominal pressure and lead to physiological changes that cause PFM weakness, resulting in bladder neck and urethral mobility, and eventually to urethral sphincter incompetence3. The PFMs are involved in urethral sphincter contraction and relaxation. During relaxation, the PFMs help to withstand pressure from the pelvic organs. As the intra-abdominal pressure increases, the PFMs contract against the pelvic organs. Therefore, sufficient contractile ability of the PFMs is pivotal to prevent urinary or fecal incontinence during relaxation and allow both micturition and defecation3,26.\n\nThis study observed a relationship between PFM strength and SUI during late pregnancy. We used a perineometer to evaluate manual strength of the PFMs because it is a fast, simple, minimally invasive, inexpensive, and highly reliable method compared with other approaches. The perineometer appears to be an appropriate objective method for measuring PFM strength and endurance, and it correlates well with those of ultrasound assessment14,27. Therefore, we used only one diagnostic device.\n\nThis article adds important findings in terms of establishing cutoff points according to PFM strength and estimated fetal weight (EFW) to predict the risk of SUI during the late third trimester in Indonesian women. PFM strength and EFW were the strongest predictors of SUI in the third trimester of pregnancy. A combined cutoff of PFM strength ≤25.5 cmH2O and estimated fetal weight ≥3,100 g was associated with a 75.39% prevalence of SUI. To our knowledge, no previous study has objectively investigated the cutoffs for these two variables during the last term of pregnancy. For instance, Baracho et al.28 used the PFM strength and birth weight obtained after delivery and found a 17% prevalence of SUI. In Indonesia, Santoso29 reported a cutoff birth weight of ≥3,325 g, but this study obtained the data after delivery as opposed to during pregnancy. We were able to identify meaningful cutoff points during pregnancy that classified these participants according to the outcomes of SUI and non-SUI.\n\nTaken together, these findings provide information that will help to guide services and health policies about the factors that predict SUI during pregnancy. This information may be useful for physicians trying to organize programs for the prevention and treatment of UI/SUI during and after pregnancy such as physiotherapy and PFMEs. PFM training is important for the prevention of SUI because it is effective in reducing urethral excursion and enhancing urethral sphincter function30. Further longitudinal studies are needed for more complete testing of this data and to investigate whether these findings can be generalized to other settings.\n\nAs a cross-sectional study, this research is limited by the lack of assessment of the mechanisms underlying the relationship between risk factors and SUI. In addition, we did not check baseline PFM strength before pregnancy or analyze data according to ethnicity. Differences in ethnicity may explain some of the differences between the findings of our study and those of others. The causal relationships between EFW, PFM, and SUI should be assessed further in future studies.\n\nIn conclusion, the two predictors of SUI identified here are modifiable by preventive approaches, mainly with physical therapy and education. Increasing PFM strength should help to prevent SUI during pregnancy. Indonesian women with PFM strength ≤25.5 cmH2O and an estimated fetal weight ≥3,100 should be encouraged to increase their PFM strength to avoid SUI late in pregnancy. We found that in women with an EFW ≥3,100 g, the risk of SUI decreased from 75.39% to 54.9% when the PFM strength increased from ≤25.5 cmH2O to >25.5 cmH2O.\n\n\nData availability\n\nHarvard Dataverse: Stress urinary incontinence in relation to pelvic floor muscle strength and its associated factors in the third trimester of pregnancy. https://doi.org/10.7910/DVN/ETXVME16.\n\nFile ‘SUI in pregnancy Data 2’ contains all underlying data analyzed in this study.\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "Acknowledgments\n\nThe authors thank to Faculty of Medicine Universitas Indonesia, all doctors and research assistants who participated in this study, and all urogynecology consultants who helped with this project.\n\n\nReferences\n\nAl-Mehaisen LM, Al-Kuran O, Lataifeh IM, et al.: Prevalence and frequency of severity of urinary incontinence symptoms in late pregnancy: a prospective study in the north of Jordan. Arch Gynecol Obstet. 2009; 279(4): 499–503. PubMed Abstract \| Publisher Full Text\n\nScarpa KP, Herrmann V, Palma PC, et al.: Prevalence and correlates of stress urinary incontinence during pregnancy: a survey at UNICAMP Medical School, São Paulo, Brazil. Int Urogynecol J Pelvic Floor Dysfunct. 2006; 17(3): 219–23. PubMed Abstract \| Publisher Full Text\n\nSangsawang B, Sangsawang N: Stress urinary incontinence in pregnant women: a review of prevalence, pathophysiology, and treatment. Int Urogynecol J. 2013; 24(6): 901–12. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOliveira Cd, Seleme M, Cansi PF, et al.: Urinary incontinence in pregnant women and its relation with socio-demographic variables and quality of life. Rev Assoc Med Bras (1992). 2013; 59(5): 460–6. PubMed Abstract \| Publisher Full Text\n\nSangsawang B: Risk factors for the development of stress urinary incontinence during pregnancy in primigravidae: a review of the literature. Eur J Obstet Gynecol Reprod Biol. 2014; 178: 27–34. PubMed Abstract \| Publisher Full Text\n\nHerbruck LF: Urinary incontinence in the childbearing woman. Urol Nurs. 2008; 28(3): 163–71; quiz 72. PubMed Abstract\n\nVan Geelen H, Ostergard D, Sand P: A review of the impact of pregnancy and childbirth on pelvic floor function as assessed by objective measurement techniques. Int Urogynecol J. 2018; 29(3): 327–338. PubMed Abstract \| Publisher Full Text\n\nSimeonova Z, Milsom I, Kullendorff AM, et al.: The prevalence of urinary incontinence and its influence on the quality of life in women from an urban Swedish population. Acta Obstet Gynecol Scand. 1999; 78(6): 546–51. PubMed Abstract\n\nHay-Smith J, Mørkved S, Fairbrother KA, et al.: Pelvic floor muscle training for prevention and treatment of urinary and faecal incontinence in antenatal and postnatal women. Cochrane Database Syst Rev. 2008; (4): CD007471. PubMed Abstract \| Publisher Full Text\n\nBradley CS, Rahn DD, Nygaard IE, et al.: The questionnaire for urinary incontinence diagnosis (QUID): validity and responsiveness to change in women undergoing non-surgical therapies for treatment of stress predominant urinary incontinence. Neurourol Urodyn. 2010; 29(5): 727–34. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMiller JM, Perucchini D, Carchidi LT, et al.: Pelvic floor muscle contraction during a cough and decreased vesical neck mobility. Obstet Gynecol. 2001; 97(2): 255–60. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNewman DK, Perucchini D: Clinical Evaluation of the Pelvic Floor Muscles. In: Baessler K, Schussler B, Burgio KL, Moore KH, Norton PA, Stanton SL, editors. ed PFR-eS, editor: London: Springer; 2008; 91–104. Publisher Full Text\n\nIsherwood PJ, Rane A: Comparative assessment of pelvic floor strength using a perineometer and digital examination. BJOG. 2000; 107(8): 1007–11. PubMed Abstract \| Publisher Full Text\n\nRahmani N, Mohseni-Bandpei MA: Application of perineometer in the assessment of pelvic floor muscle strength and endurance: a reliability study. J Bodyw Mov Ther. 2011; 15(2): 209–14. PubMed Abstract \| Publisher Full Text\n\nUyar Y, Baytur YB, Inceboz U: Perineometer and digital examination for assessment of pelvic floor strength. Int J Gynaecol Obstet. 2007; 98(1): 64–5. PubMed Abstract \| Publisher Full Text\n\nYunita A: Stress urinary incontinence in relation to pelvic floor muscle strength and its associated factors in the third trimester of pregnancy. Harvard Dataverse, V3. 2019. http://www.doi.org/10.7910/DVN/ETXVME\n\nWhitford HM, Alder B, Jones M: A cross-sectional study of knowledge and practice of pelvic floor exercises during pregnancy and associated symptoms of stress urinary incontinence in North-East Scotland. Midwifery. 2007; 23(2): 204–17. PubMed Abstract \| Publisher Full Text\n\nMartins G, Soler ZA, Cordeiro JA, et al.: Prevalence and risk factors for urinary incontinence in healthy pregnant Brazilian women. Int Urogynecol J. 2010; 21(10): 1271–7. PubMed Abstract \| Publisher Full Text\n\nDinc A: Prevalence of Urinary Incontinence During Pregnancy and Associated Risk Factors. Low Urin Tract Symptoms. 2018; 10(3): 303–307. PubMed Abstract \| Publisher Full Text\n\nPandit M, DeLancey JO, Ashton-Miller JA, et al.: Quantification of intramuscular nerves within the female striated urogenital sphincter muscle. Obstet Gynecol. 2000; 95(6 Pt 1): 797–800. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPerucchini D, DeLancey JO, Ashton-Miller JA, et al.: Age effects on urethral striated muscle. I. Changes in number and diameter of striated muscle fibers in the ventral urethra. Am J Obstet Gynecol. 2002; 186(3): 351–5. PubMed Abstract \| Publisher Full Text\n\nRud T: Urethral pressure profile in continent women from childhood to old age. Acta Obstet Gynecol Scand. 1980; 59(4): 331–5. PubMed Abstract \| Publisher Full Text\n\nThomason AD, Miller JM, Delancey JO: Urinary incontinence symptoms during and after pregnancy in continent and incontinent primiparas. Int Urogynecol J Pelvic Floor Dysfunct. 2007; 18(2): 147–51. PubMed Abstract \| Publisher Full Text\n\nFritel X, Fauconnier A, Bader G, et al.: Diagnosis and management of adult female stress urinary incontinence: guidelines for clinical practice from the French College of Gynaecologists and Obstetricians. Eur J Obstet Gynecol Reprod Biol. 2010; 151(1): 14–9. PubMed Abstract \| Publisher Full Text\n\nChin HY, Chen MC, Liu YH, et al.: Postpartum urinary incontinence: a comparison of vaginal delivery, elective, and emergent cesarean section. Int Urogynecol J Pelvic Floor Dysfunct. 2006; 17(6): 631–5. PubMed Abstract \| Publisher Full Text\n\nKatarina PI, Ljiljana MS, Aleksandra JS: The influence of various risk factors on the strength of pelvic floor muscle in women Uticaj razli č itih faktora rizika na ja č inu miši ć a poda karlice kod žene. 2017; 74(6): 557–63. Publisher Full Text\n\nvan Veelen A, Schweitzer K, van der Vaart H: Ultrasound assessment of urethral support in women with stress urinary incontinence during and after first pregnancy. Obstet Gynecol. 2014; 124(2 Pt 1): 249–56. PubMed Abstract \| Publisher Full Text\n\nBaracho SM, Barbosa da Silva L, Baracho E, et al.: Pelvic floor muscle strength predicts stress urinary incontinence in primiparous women after vaginal delivery. Int Urogynecol J. 2012; 23(7): 899–906. PubMed Abstract \| Publisher Full Text\n\nSantoso IB: Budi Iman Santoso Assessment (BISA): a model for predicting levator ani injury after vaginal delivery. Medical Journal Indonesia. 2012; 21(2): 102–7. Publisher Full Text\n\nMcLean L, Varette K, Gentilcore-Saulnier E, et al.: Pelvic floor muscle training in women with stress urinary incontinence causes hypertrophy of the urethral sphincters and reduces bladder neck mobility during coughing. Neurourol Urodyn. 2013; 32(8): 1096–102. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "54283", "date": "21 Oct 2019", "name": "Jimmy Nomura", "expertise": [ "Reviewer Expertise Urogynecology", "voiding dysfunction in women" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe present article is an interesting study. From the background of the study, showed that stress urinary incontinence (SUI) is a common in pregnancy caused by several things including uterine growth, hormonal influences, and biomechanical disorders in the pelvis. These factors make the muscle tone and muscle strengh in the pelvic region decrease during pregnancy.\nThe authors describe that the background of the research concisely and is easy to understand the reasons for doing this research. Previously, a lot of studies have been done to investigate the effect of pregnancy on SUI. but little evidence has shown about PFM cut off value for SUI. The authors investigate the relationships between SUI and obstetric factors during pregnancy and patient characteristics in Indonesian women, to identify the clinical predictors of SUI, particularly PFM strength, in the third trimester of pregnancy using a digital examination perineometer, and to develop a predictive model of SUI with cutoff points for use in practice for the early detection, and possible prevention, of modifiable risk factors that contribute to the development of SUI during pregnancy. From the present study, we can see the the cut off PFM strengh and fetal weight to correlated with SUI in pregnancy.\nI think that the present study does not have big problem and is well-organized.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "65941", "date": "03 Jul 2020", "name": "Angie Rantell", "expertise": [ "Reviewer Expertise Urogynaecology", "pelvic floor dysfunction / sexual dysfuntion" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI think that this is a sound idea but I have several questions about the quality of the study and paper. In the introduction the definition of SUI is not correct, they have provided the definition of UI. They comment about prevalence in pregnancy but not about how many it resolves in post partum.\nMethods\nThere is no standardization of the cough test - how was a full bladder defined.\n\nWhy was PFMT taught by Dr's - is this normal practice, were physios involved.\n\nMode of delivery - what about forceps - do you not do this?\nDiscussion\nThere is no mention about what is normal PFM strength on perineometry in women without SUI who are not pregnant.\n\nHow will this going to change practice?\n\nYou report that you found modifiable factors but how is EFW modifiable.\n\nThere is no comment about what is normal antenatal care regarding PFMT and advice.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-1684
https://f1000research.com/articles/8-262/v1	06 Mar 19	{ "type": "Research Article", "title": "Diversity and Molecular Characterization of Mosquitoes (Diptera: Culicidae) in selected ecological regions in Kenya.", "authors": [ "Moni Makanda", "Gladys Kemunto", "Lucy Wamuyu", "Joel Bargul", "Jackson Muema", "James Mutunga", "Gladys Kemunto", "Lucy Wamuyu", "Joel Bargul", "Jackson Muema", "James Mutunga" ], "abstract": "Mosquitoes play a predominant role as leading agents in the spread of vector-borne diseases and consequent mortality in humans. Despite reports on increase of new and recurrent mosquito borne-disease outbreaks such as chikungunya, dengue fever and Rift valley fever in Kenya little is known about the genetic characteristics and diversity of the vector species that have been incriminated in transmission of disease pathogens. In this study, we identified mosquito species across Kisumu, Kilifi and Nairobi Counties and determined their genetic diversity and phylogenetic relationships. PCR was used to amplify and sequence the partial cytochrome oxidase subunit 1 (CO1) gene of mosquito samples. Molecular-genetic and phylogenetic analysis of the partial cytochrome oxidase subunit 1 (CO1) gene was employed to identify their relationships with known mosquito species. Fourteen (14) haplotypes belonging to genus Aedes, nine (9) haplotypes belonging to genus Anopheles and twelve (12) haplotypes belonging to genus Culex were identified in this study. Findings from this study revealed a potentially new haplotype belonging to Anopheles genus and reported the first molecular characterization of Aedes cummnisii in Kenya. Sequence results revealed variation in mosquito species from Kilifi, Kisumu and Nairobi. Since vector competence varies greatly across species and species-complexes and is strongly associated with specific behavioural adaptations, proper species identification is important for vector control programs.", "keywords": [ "Aedes", "Anopheles", "Culex", "chikungunya", "Rift valley fever", "dengue fever" ], "content": "Introduction\n\nMosquitoes are vectors responsible for transmission of numerous pathogens causing diseases such as; malaria, lymphatic filariases, avian malaria and arboviruses such as Dengue virus, Chikungunya virus, Yellow fever, West Nile Fever, and Zika virus (Charrel et al., 2007; Semenza, 2014). Africa is one of the major hosts of mosquitoes responsible for mosquito-borne viruses (Braack et al., 2018) that are of great medical importance and contribute to the current global public health threat (Enserink, 2007; Gubler, 2002; Higgs, 2014). Seasonal and environmental changes play a role in the global distribution of mosquito species and the arboviruses they transmit (Anyamba et al., 2001; Hasnan et al., 2016). The global spread of vector-borne diseases has resulted to multiple calls on nations to enhance surveillance of emerging arboviruses that requires understanding of the species composition and distribution of potential mosquito vectors (Grout et al., 2017; Kollars et al., 2016).\n\nIn the recent past, there has been an increasing spread of mosquito-borne viruses such as Chikungunya, dengue fever and Rift Valley fever in Kenya, thus prompting a need for further research (Johnson et al., 1982; Konongoi et al., 2018). The available literature on mosquitoes in Kenya mainly addresses aspects of morphological identification of mosquito vectors and limited molecular characterization (Lutomiah et al., 2013; Mwangangi et al., 2013). Despite mosquitoes being a major public health challenge in Kenya, little is known about their species diversity and distribution along different ecological zones such as the Kenyan coast and Kenya’s capital city. Subsequently, population genetic studies on mosquito vectors in Kenya have focused on the Anopheles genus because of its importance in endemic malaria transmission (Chen et al., 2006; Chen et al., 2004; Lukindu et al., 2018). In addition, most of the studies on mosquito vector composition and diversity are based on mosquitoes confined to a single habitat or with a limited habitat range (Ajamma et al., 2016a; Muturi et al., 2006). The species composition and distribution of Anopheline mosquitoes in Kenya, particularly along the Kenyan coast, have broadly been reported over that of Culcine mosquitoes (Mbogo et al., 2003; Midega et al., 2007, Midega et al., 2010). Moreover, little has been documented on the species composition and diversity of all mosquito groups by use of molecular markers. As such, understanding the species composition and diversity patterns of the suggested vectors is pivotal to the judicious deployment of existing vector control strategies and the development of new effective vector control interventions (Kraemer et al., 2016).\n\nIn this study, we employed molecular genetic techniques, involving PCR and sequencing of cytochrome oxidase subunit 1 (CO1) gene to identify and characterize mosquito species in Nairobi, Kisumu and Kilifi Counties in Kenya.\n\n\nMethods\n\nThis study was carried out at Nairobi, Kilifi and Kisumu Counties in Kenya. Kilifi and Kisumu regions were chosen purposively due to their high abundance of mosquito vectors (WHO, 2017) and vector-borne disease burden, while Nairobi region was selected because it’s a major international and domestic destination for both humans and parasites (Wesolowski et al., 2012). Two sampling sites were randomly selected from each of the three regions as follows: Kisumu; Ahero and Kisumu town, Kilifi; Kilifi town and Mazingira Park and Nairobi; Nairobi city centre, and Northern Bypass (Figure 1).\n\nThe trapping of mosquitoes was carried out in the respective counties during the dry season (January–February 2018) and wet season (March–April 2018). The captures were conducted day and night using the Pyrethrum Spray Catch (PSC) method (Ndiath et al., 2011). The specimens were adult mosquitoes, which were morphologically sorted in the field into their respective genera, and transported in liquid nitrogen to the laboratory for further molecular analysis. A total of 2,438 adult mosquitoes were collected. Of these, 894, 824 and 720 adult mosquito samples were collected in Nairobi, Kisumu and Kilifi respectively. From the overall collection, 300 hundred mosquitoes per county were randomly selected for PCR. A total of 25 sequences per study region were used for phylogenetic and genetic diversity analysis (Hale et al., 2012).\n\nTotal genomic DNA was extracted from whole body of individual mosquitoes using the Collins’ protocol (Collins et al., 1987) with minor modifications. A DNA homogenizing buffer (containing 0.1 M NaCl, 0.2 M sucrose, 0.01 M EDTA and 0.03 M Tris pH 8) was mixed with a lysis buffer (containing 0.25 M EDTA, 2.5% w/v SDS and 0.5 M Tris pH 9.2) in the ratio 4:1 to make up the grinding buffer (GB). Each mosquito was homogenized in 100 цl of the GB, using a hand-held pestle homogenizer and incubated for 30 min at 55°C. Into each sample, 14 цl of 8 M potassium acetate (KAc), a deproteinating reagent was added and then incubated for 30 min at room temperature before centrifuging at 13,000 rpm for 15 min to get the supernatant that contains the nucleic acid component. Ethanol precipitation 95% was used to precipitate the genomic DNA. Centrifugation at 13,000 rpm for 10 minutes to obtain a pellet that is the nucleic acid. Followed by a washing step using 70% ethanol. The DNA pellet was suspended in 100 μL of T.E buffer pH 7.2 and stored at –20°C awaiting subsequent experimental procedures.\n\nThe primer set ;Forward (LCO1490_GGTCAACAAATCATAAAGATATTGG) and Reverse (HCO2198 TAAACTTCAGGGTGACCAAAAAATCA) synthesized by Macrogen (OG180803-187) and previously published by Folmer et al. were used in molecular identification of the mosquito species (Folmer et al., 1994). In a 10 µL PCR reaction volume, the PCR mix consisted of 2 µL 1× HOT FIREPol® Eva Green mix (Solis BioDyne, Tartu, Estonia) catalogue number 08-31-00008, 6 µL of nuclease-free water, 0.5 picomoles of each primer and 1 µL of the DNA template. The fragments were amplified using applied biosystems ProFlex SN 297802057 thermocycler under the following cycling parameters; initial denaturation for 15 min at 95°C, followed by 35 cycles of denaturation at 95°C for 30 sec, annealing at 50°C (Anopheles, Aedes, Culex) for 30 sec, and extension at 72°C for 30 sec, and a final extension at 72°C for 7 min. The PCR products from the amplification of the mitochondrial cytochrome c oxidase 1 (CO1) region of the mosquito after purification using QIAquick® gel extraction kit catalogue number 28706, were shipped for sequencing at Macrogen Inc., South Korea.\n\nResultant mitochondrial cytochrome c oxidase 1 (CO1) sequence chromatograms were edited and visualized using Chromas Lite version 2.6.5. The sequences were deposited in GenBank and accession numbers assigned accordingly. Consensus sequences were aligned using ClustalX version 2 (Thompson et al., 1997), and visualized using Seaview Version 4.7 (Gouy et al., 2010). Unique sequences (haplotypes) were identified using DnaSP version 6 (Librado & Rozas, 2009). Sequence polymorphisms were identified using DnaSP and visualized using Jalview version 2.10.5 (Waterhouse et al., 2009). DNA sequence divergence was analysed using DnaSP. These unique sequences were compared with reference sequences from other parts of the world, selected to represent the Aedes, Anopheles and Culex genera previously reported and available from GenBank (Benson et al., 2011). Other sequences similar to the study sequences in GenBank obtained using the Blastn algorithm were also included in the analysis. Multiple alignment and comparison of the study sequences and GenBank references were performed using ClustalX. Phylogenetic and molecular evolutionary analyses were conducted using Software for Molecular Evolutionary Genetics (MEGA7) (Kumar et al., 2016). Phylogenetic trees were constructed using the maximum likelihood (ML) method rooted using Lutzomyia longipalpis. The phylogenetic trees were estimated using the best-fit general time-reversible (GTR) model of nucleotide substitution with gamma-distributed rate variation among sites. Bootstrap resampling process (1000 replications) was employed to assess the robustness of individual nodes of phylogeny (only >50% were indicated). The resultant tree was visualized using Dendroscope version 3 (Huson & Scornavacca, 2012).\n\n\nResults\n\nFrom each study site 25 CO1 gene amplicons were sequenced for phylogenetic analysis. In total, 14 haplotypes belonging to genera Aedes, 9 haplotypes belonging to genera Anopheles and 12 haplotypes belonging to genera Culex were identified through CO1 sequence analysis; sequences were deposited in GenBank and assigned accession numbers (Table 1). Sequence analysis revealed a unique Anopheles haplotype (GenBank accession number, MK300230) (Figure 3). Subsequently, haplotypes of Anopheles gambiae, Anopheles funestus, Aedes cumminsii, Aedes aegypti, Culex pipiens and Culex sitiens were found to be distributed across Kilifi, Kisumu and Nairobi mosquito populations (Table 1).\n\nDiversity indices for the three populations, based on sequenced results were calculated as shown in (Table 2). Average number nucleotide differences (k), nucleotide diversity Pi (π) and haplotype diversity (Hd) varied among the species (Table 2).\n\n2Hap: number of haplotypes; S: number of polymorphic Segregating Sites; k: The average number of nucleotide differences; Pi (π): nucleotide diversity; Hd: haplotype gene diversity.\n\nPhylogenetic analysis of fourteen (14) Aedes haplotypes from Kilifi and Nairobi with similar sequences based on Blastn (NCBI) search and sequences of known Aedes identity revealed that study Aedes haplotype Accession number MK300225 clustered with Aedes cumminsii (Figure 2). Study haplotypes Accession number; MK300216, MK300217, MK300218, MK300219, MK300220, MK300221, MK300222, MK300223, MK300224, MK300226, MK300227, MK300228 and MK300229 clustered with Aedes aegypti that has been previously identified in France (Accession number HQ688297.1). Significantly, they also clustered with Aedes aegypti (Accession number KX420485.1, KX420429.1 and KU380400.1) previously reported in Nyanza-Kisumu, Kenya (Figure 2). Genetic divergence between study Aedes haplotypes identified in Kilifi and Nairobi and Aedes species they clustered with (sequences of known species obtained from GenBank) was variable (Table 3). There was limited divergence between Aedes aegypti (Accession number KX420485) that has previously been identified in Nyanza-Kisumu, Kenya and study haplotypes MK300216, MK300222, MK300218 and MK300221. Aedes aegypti (Accession number KU380400.1) that has been reported in Nyanza-Kisumu, Kenya before showed limited divergence with study haplotype MK300217. Limited divergence was also identified between haplotype MK300224 and Aedes aegypti (Accession number HQ688297.1) that has been characterized in France. Greater divergence and heterogeneity was observed between Aedes aegypti and study haplotypes MK300225, MK300219 and MK300229. Study haplotypes MK300216, MK300220, MK300223, MK300227 and MK300228 formed a distinct clade with other Aedes aegypti of known identity (Figure 2).\n\nThe scale represents the number of differences between sequences (0.02=2%). The gamma correction for rate heterogeneity was 0.1963. The analysis involved 46 nucleotide sequences. There were a total of 657 positions in the final dataset.\n\nPhylogenetic analysis of haplotypes with similar sequences to those of known identity showed a clustering of study Anopheles haplotype MK300231 and MK300232 with Anopheles funestus. Notably, they also clustered with Anopheles funestus (Accession number MH299888.1 and KU380404.1) that has been reported in Kilifi and Baringo counties in Kenya respectively (Figure 3). Study haplotype MK300233, MK300234, MK300235, MK300236, MK300237 and MK300238 clustered with Anopheles gambiae that have previously been isolated in Uganda (Accession number MG753695.1, MG753730.1 and MG753745.1) (Figure 3). Anopheles haplotype MK300230 formed its own distinct clade. This study haplotype MK300230 may be a new species or novel haplotype not yet described (Figure 3). Genetic divergence between Anopheles haplotypes identified Kisumu, Kilifi, Nairobi and Anopheles species they clustered with was variable in some haplotypes while others were not variable (Table 4). There was very limited divergence and heterogeneity between Anopheles funestus and study haplotype MK300231 and MK300232. There was no divergence between Anopheles gambiae (Accession number DQ792577.1 and MG753695.1) and study haplotype MK300234. Anopheles gambiae (Accession number MG753695.1) has been identified in Uganda before. Study haplotypes MK300235 MK300233 MK300238 MK300236 and MK300237 showed limited divergence with Anopheles gambiae.\n\nThe gamma correction for rate heterogeneity was 0.1647. The analysis involved 57 nucleotide sequences. There were a total of 658 positions in the final dataset.\n\nFrom the phylogenetic analysis, we further established that 12 Culex haplotypes from Kilifi, Kisumu and Nairobi, and similar sequences of known identity based on Blastn (NCBI) showed a clustering of study haplotype MK300240, MK300242, MK300246, MK300247, MK300248, MK300249 and MK300250 with Culex pipiens that have been identified in different regions of the world. Importantly, they clustered with Culex pipiens that has previously been identified in Nyanza-Kisumu, Kenya (Accession number KU380381.1, KU380372.1) (Figure 4). Study haplotypes MK300239, MK300241, MK300243, MK300244, MK300245 clustered with Culex sitiens that has been previously identified in Australia (Accession number MG712559.1) (Figure 4). Genetic divergence between Culex haplotypes identified Kisumu, Kilifi, Nairobi and reference Culex species was slightly variable in some species, while other species showed no divergence (Table 5).\n\nThe gamma correction for rate heterogeneity was is 0.1790. The analysis involved 62 nucleotide sequences. There were a total of 658 positions in the final dataset.\n\n\nDiscussion\n\nThis study identified Aedes aegypti in both Kilifi and Nairobi populations and Aedes cummnisii in the Kilifi population only. Anopheles gambiae was identified in both Kisumu and Nairobi population whereas Anopheles funestus was identified in Kisumu population only. A potentially novel Anopheles haplotype MK300230 was identified in Kilifi population. Culex pipiens was identified in all the three populations; Kisumu, Nairobi and Kilifi while Culex sitiens was only identified in the Kilifi population. The greatest diversity was in the genus Aedes that has 14 haplotypes, followed by Culex 12 and Anopheles nine, this is consistent with other studies looking at mosquito diversity in different ecological regions in Kenya (Mwangangi et al., 2012). Similarly, out of the 35 mosquitoes haplotypes identified in Kilifi, Nairobi and Kisumu regions, one Culex haplotype MK300242 from this study has been previously reported at Kisumu-Nyanza in Kenya and in Portugal (Ajamma et al., 2016b; Mixão et al., 2016), and one Anopheles haplotype MK300234 in Uganda (Lukindu et al., 2018). The Kilifi mosquito population had the greatest diversity and abundance of mosquito species, possibly due to its geographical position, human activities and natural climatic conditions.\n\nAedes cummnisii has been morphologically identified in Kenya before (Mwangangi et al., 2006), however, this study reports the first molecular characterization of Aedes cummnisii in Kenya. Aedes haplotypes between Kilifi and Nairobi populations were divergent based on nucleotide diversity tests; this could be due to different climatic zones. Thus, diversity in vector haplotypes plays an important role in vector control and management practices and epidemiology of vector borne diseases (Murugan et al., 2016). Phylogenetic analysis showed presence of two Aedes species that is Aedes cummnisii and Aedes aegypti, in Kilifi, while Nairobi had only Aedes aegypti (Figure 2 and Table 1). This study has identified 4 different Aedes aegypti haplotypes in Nairobi. Previous studies have indicated presence of only a few Aedes aegypti in Nairobi (Kinuthia et al., 2017). There is therefore an increase in diversity in Aedes aegypti species from Nairobi; diversity and spread of Aedes aegypti has been associated with expansion on arboviral infection (Woolhouse et al., 1997). The diversity of Aedes aegypti in Nairobi could be the result of high population density (Gubler & Clark, 1995), poor sanitation and waste disposal as well as water management (Monath, 1994). The Kilifi population had genetically diverse forms of Aedes aegypti (Table 2). Aedes aegypti is widespread on the Kenyan coast (McDonald, 1977; Teesdale, 1955). It is the principal vector of dengue virus, chikungunya, and urban yellow fever virus (Reiter, 2010), and was predominated in the Kilifi samples. This may contribute to the high susceptibility to dengue-outbreak reported in the region (Baba et al., 2016; Chepkorir et al., 2014). Secondly, factors relating to availability of breeding sites, temperature or altitudinal differences may have influenced the diversity patterns of Aedes aegypti in Kilifi (Barrera et al., 2011). Evidence of high diversity of Aedes aegypti in Kilifi also means that the Kenyan coast is consistently at higher risk of Yellow Fever transmission (Agha et al., 2017). Kilifi lies in between Malindi and Mombasa cities which are popular destinations for international tourism as well as maritime industry, and where Aedes aegypti is widespread (Ngugi et al., 2017). Human trade and travel may bolster movement of Aedes aegypti (Powell & Tabachnick, 2013) and contribute to diversity of the species, in addition invasion risk related to human travel has become far more severe (Egizi et al., 2016; Wilder-Smith & Gubler, 2008). Phylogenetic relationship between Aedes species from this study and other Aedes species of known identity from GenBank showed clustering with Aedes cummnisii and Aedes aegypti at a high bootstrap value (>90%) at the defining node on the phylogenetic tree (Figure 2). However, genetic diversity between Aedes species from this study and those of known identity from GenBank was variable (Table 3).\n\nAnopheles species were distributed across the three study populations Kisumu, Nairobi and Kilifi (Table 1). Anopheles species between Kilifi, Kisumu and Nairobi populations were highly divergent as analyzed using molecular markers. Nairobi had only one haplotype of Anopheles gambiae (Table 1). Anopheles mosquitoes have also been reported in places where malaria has been eradicated and also in malaria non endemic regions thus increasing the risk of reintroduction of malaria as well as spreading of malaria to new areas (Martens & Hall, 2000). Other than transmitting Malaria, Anopheles mosquitoes have been indicated as a carriers of arboviruses including West Nile Virus and Japanese Encephalitis (Thenmozhi et al., 2006), as well as viruses that cause O’nyong-nyong and Chikungunya fevers (Vanlandingham et al., 2005). This study has indicated high diversity of Anopheles haplotypes in the Kisumu population, having detected Anopheles gambiae and Anopheles funestus (Table 2). High diversity of Anopheles vector is a key feature for consideration in Anopheles management and has been associated with the rise in malaria transmission (Loaiza et al., 2012). The low diversity of Anopheles species in Kilifi and Nairobi may be attributed to the Great Rift Valley, high-elevation mountains in western Kenya. The vast arid area east of the Great Rift Valley inhibits human settlement, thus restricting Anopheles funestus gene flow between coastal and western Kenya (Lukindu et al., 2018). Anopheles funestus is closely associated with human dwellings and therefore plays an important role in the transmission of Malaria (Kweka et al., 2013). Anopheles gambiae haplotypes in Kisumu were diverse, this is consistent with other studies that have reported a high genetic diversity of Anopheles gambiae in Kisumu Kenya (Chen et al., 2004). Phylogenetic analysis (Figure 3) and nucleotide diversity tests (Table 4) showed no divergence between Kisumu Anopheles gambiae haplotype MK300234 with Anopheles gambiae MG753695.1, used as reference that was previously isolated in Uganda (Lukindu et al., 2018). This indicates the presence of genetically identical Anopheles gambiae between Kenya and Uganda which could be attributed to cross-border migration across Lake Victoria. Therefore, this could suggest that, these species share the same ecological niche or ancestral divergence. Anopheles gambiae (s.s.) (formerly Anopheles gambiae S-form) is a main vector of malaria in sub-Saharan Africa, where 91% of an estimated 445,000 malaria deaths worldwide occurred in 2016 (CDC - Malaria - About Malaria - Disease). Presence of both Anopheles gambiae and Anopheles funestus in Kisumu suggest that the area is still at high risk of malaria transmission. This study has identified a potentially new haplotype of Anopheles species MK300230 in Kilifi (Figure 3). Through molecular techniques new haplotypes of Anopheles species are continually being identified; for instance, new species of Anopheles nuneztovari have been identified in Brazil (Scarpassa et al., 2016).\n\nCulex pipiens was distributed across Kilifi, Kisumu and Nairobi population while Culex sitiens was only identified in Kilifi population (Table 1). Culicidae is a large and abundant group that occurs throughout temperate and tropical regions of the world, as well as the peri Arctic Circle (Schäfer & Lundström, 2001). Culex mosquitos are an important vector of zoonotic infection Filariasis. Human filariasis infection is a major public health concern. Approximately 66% of those at risk of infection is in the South-East Asia Region and 33% in the African Region (“WHO \| World malaria report 2017,” 2018). Although Culex pipiens is ornithophilic it can also feed on humans and mammals (Reisen et al., 1990; Reisen & Reeves, 1990) and thus capable to transmit West Nile Virus to Humans. Culex pipiens (Linnaeus) has been identified as the primary vector of West Nile virus (Turell et al., 2000). Kenyan strain of Culex pipiens has been confirmed to be capable of transmitting West Nile virus and its circulation among humans in Kenya has been detected (Lutomiah et al., 2011; Morrill et al., 1991). Therefore, the distribution of Culex pipiens across Kilifi, Nairobi and Kisumu could increase the risk of West Nile virus transmissions/outbreaks in most parts of Kenya. Culex pipiens haplotype MK300242 was identified in both Kilifi and Kisumu population (Figure 4). This study reports distribution of identical mosquito vector species between populations. Phylogenetic analysis revealed Culex pipiens haplotype MK300242 from this study showed no divergence to the Culex pipiens sequences LC102132.1 from Portugal and KU380381.1, KU380372.1 from Nyanza Kenya (Table 5). This study identified Culex sitiens in the Kilifi population only, Culex sitiens has been found to tolerate saline waters, in Oman it has been successfully isolated from brackish water (Roberts, 1996). Consequently, parasites such as Microsporidium, Amblyospora have been isolated from Culex sitiens mosquito in Coastal Kenya (Sabwa et al., 1984).\n\n\nConclusion\n\nThis study has identified mosquito vectors that could spread arboviral pathogens distributed across Kilifi, Nairobi and Kisumu counties. The distribution varies in density where in some cases vector distribution is limited to particular areas which could be attributed to ecological and environmental variations.\n\n\nData availability\n\nCulicidae cytochrome c oxidase subunit 1 (COI) gene, partial cds; mitochondrial. PopSet 1573759763: https://www.ncbi.nlm.nih.gov/popset/1573759763?report=genbank. Accession numbers MK300216 – MK300250", "appendix": "Grant information\n\nJB is a Training Health Researchers into Vocational Excellence 2 (THRiVE-2) fellow, a program of the Developing Excellence in Leadership, Training and Science (DELTAS) Africa Initiative [DEL-15-011]. The DELTAS Africa Initiative is an independent funding scheme of the African Academy of Sciences (AAS)’s Alliance for Accelerating Excellence in Science in Africa (AESA) and supported by the New Partnership for Africa’s Development Planning and Coordinating Agency (NEPAD Agency) with funding from the Wellcome Trust [107742/Z/15/Z] and the UK government. The views expressed in this publication are those of the author(s) and not necessarily those of AAS, NEPAD Agency, Wellcome Trust or the UK government.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe thank International Centre of Insect Physiology and Ecology (ICIPE) for hosting the research work.\n\n\nReferences\n\nAgha SB, Tchouassi DP, Bastos ADS, et al.: Dengue and yellow fever virus vectors: seasonal abundance, diversity and resting preferences in three Kenyan cities. Parasit Vectors. 2017; 10(1): 628. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAjamma YU, Mararo E, Omondi D, et al.: Rapid and high throughput molecular identification of diverse mosquito species by high resolution melting analysis. [version 1; referees: 2 approved]. F1000Res. 2016a; 5: 1949. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAjamma YU, Villinger J, Omondi D, et al.: Composition and Genetic Diversity of Mosquitoes (Diptera: Culicidae) on Islands and Mainland Shores of Kenya's Lakes Victoria and Baringo. J Med Entomol. 2016b; 53(6): 1348–1363. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAnyamba A, Linthicum KJ, Tucker CJ: Climate-disease connections: Rift Valley Fever in Kenya. Cad Saude Publica. 2001; 17 Suppl: 133–140. PubMed Abstract \| Publisher Full Text\n\nBaba M, Villinger J, Masiga DK: Repetitive dengue outbreaks in East Africa: A proposed phased mitigation approach may reduce its impact. Rev Med Virol. 2016; 26(3): 183–96. PubMed Abstract \| Publisher Full Text\n\nBarrera R, Amador M, MacKay AJ: Population dynamics of Aedes aegypti and dengue as influenced by weather and human behavior in San Juan, Puerto Rico. PLoS Negl Trop Dis. 2011; 5(12): e1378. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBenson DA, Karsch-Mizrachi I, Lipman DJ, et al.: GenBank. Nucleic Acids Res. 2011; 39(Database issue): D32–D37. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBraack L, Gouveia de Almeida AP, Cornel AJ, et al.: Mosquito-borne arboviruses of African origin: review of key viruses and vectors. Parasit Vectors. 2018; 11(1): 29. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCDC - Malaria - About Malaria - Disease. [cited 2019 Mar 2]. Reference Source\n\nCharrel RN, de Lamballerie X, Raoult D: Chikungunya outbreaks--the globalization of vectorborne diseases. N Engl J Med. 2007; 356(8): 769–71. PubMed Abstract \| Publisher Full Text\n\nChen H, Fillinger U, Yan G: Oviposition behavior of female Anopheles gambiae in western Kenya inferred from microsatellite markers. Am J Trop Med Hyg. 2006; 75(2): 246–50. PubMed Abstract \| Publisher Full Text\n\nChen H, Minakawa N, Beier J, et al.: Population genetic structure of Anopheles gambiae mosquitoes on Lake Victoria islands, west Kenya. Malar J. 2004; 3: 48. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChepkorir E, Lutomiah J, Mutisya J, et al.: Vector competence of Aedes aegypti populations from Kilifi and Nairobi for dengue 2 virus and the influence of temperature. Parasit Vectors. 2014; 7: 435. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCollins FH, Mendez MA, Rasmussen MO, et al.: A ribosomal RNA gene probe differentiates member species of the Anopheles gambiae complex. Am J Trop Med Hyg. 1987; 37(1): 37–41. PubMed Abstract \| Publisher Full Text\n\nEgizi A, Kiser J, Abadam C, et al.: The hitchhiker's guide to becoming invasive: exotic mosquitoes spread across a US state by human transport not autonomous flight. Mol Ecol. 2016; 25(13): 3033–3047. PubMed Abstract \| Publisher Full Text\n\nEnserink M: Infectious diseases. Chikungunya: no longer a third world disease. Science. 2007; 318(5858): 1860–1861. PubMed Abstract \| Publisher Full Text\n\nFolmer O, Black M, Hoeh W, et al.: DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Mol Mar Biol Biotechnol. 1994; 3(5): 294–299. PubMed Abstract\n\nGouy M, Guindon S, Gascuel O: SeaView version 4: A multiplatform graphical user interface for sequence alignment and phylogenetic tree building. Mol Biol Evol. 2010; 27(2): 221–224. PubMed Abstract \| Publisher Full Text\n\nGrout A, Howard N, Coker R, et al.: Guidelines, law, and governance: disconnects in the global control of airline-associated infectious diseases. Lancet Infect Dis. 2017; 17(4): e118–e122. PubMed Abstract \| Publisher Full Text\n\nGubler DJ, Clark GG: Dengue/dengue hemorrhagic fever: the emergence of a global health problem. Emerg Infect Dis. 1995; 1(2): 55–57. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGubler DJ: Epidemic dengue/dengue hemorrhagic fever as a public health, social and economic problem in the 21st century. Trends Microbiol. 2002; 10(2): 100–103. PubMed Abstract \| Publisher Full Text\n\nHale ML, Burg TM, Steeves TE: Sampling for microsatellite-based population genetic studies: 25 to 30 individuals per population is enough to accurately estimate allele frequencies. PLoS One. 2012; 7(9): e45170. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHasnan A, Dom NC, Rosly H, et al.: Quantifying the distribution and abundance of aedes mosquitoes in dengue risk areas in shah alam, selangor. Procedia Soc Behav Sci. 2016; 234: 154–163. Publisher Full Text\n\nHiggs S: Chikungunya virus: a major emerging threat. Vector Borne Zoonotic Dis. 2014; 14(8): 535–536. PubMed Abstract \| Publisher Full Text\n\nHuson DH, Scornavacca C: Dendroscope 3: an interactive tool for rooted phylogenetic trees and networks. Syst Biol. 2012; 61(6): 1061–1067. PubMed Abstract \| Publisher Full Text\n\nJohnson BK, Ocheng D, Gichogo A, et al.: Epidemic dengue fever caused by dengue type 2 virus in Kenya: preliminary results of human virological and serological studies. East Afr Med J. 1982; 59(12): 781–784. PubMed Abstract\n\nKinuthia G, Ngure V, Kamau L, et al.: Survey of urban mosquitoes species (Diptera: Culicidae) with focus on waste water channels as larval habitats in Nairobi industrial area, Kenya. Afr J Health Sci. 2017; 30(2): 120–138. Reference Source\n\nKollars TG, Kollars P, Hulsey B: Reducing the risk to marine ports from invasive mosquito species, zika, dengue, chikungunya viruses and filariasis. Int J Mech Sci. 2016; 4(2): 70. Publisher Full Text\n\nKonongoi SL, Nyunja A, Ofula V, et al.: Human and entomologic investigations of chikungunya outbreak in Mandera, Northeastern Kenya, 2016. PLoS One. 2018; 13(10): e0205058. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKraemer MUG, Hay SI, Pigott DM, et al.: Progress and Challenges in Infectious Disease Cartography. Trends Parasitol. 2016; 32(1): 19–29. PubMed Abstract \| Publisher Full Text\n\nKumar S, Stecher G, Tamura K: MEGA7: Molecular Evolutionary Genetics Analysis Version 7.0 for Bigger Datasets. Mol Biol Evol. 2016; 33(7): 1870–1874. PubMed Abstract \| Publisher Full Text\n\nKweka EJ, Kamau L, Munga S, et al.: A first report of Anopheles funestus sibling species in western Kenya highlands. Acta Trop. 2013; 128(1): 158–161. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLibrado P, Rozas J: DnaSP v5: a software for comprehensive analysis of DNA polymorphism data. Bioinformatics. 2009; 25(11): 1451–1452. PubMed Abstract \| Publisher Full Text\n\nLoaiza JR, Bermingham E, Sanjur OI, et al.: Review of genetic diversity in malaria vectors (Culicidae: Anophelinae). Infect Genet Evol. 2012; 12(1): 1–12. PubMed Abstract \| Publisher Full Text\n\nLukindu M, Bergey CM, Wiltshire RM, et al.: Spatio-temporal genetic structure of Anopheles gambiae in the Northwestern Lake Victoria Basin, Uganda: implications for genetic control trials in malaria endemic regions. Parasit Vectors. 2018; 11(1): 246. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLutomiah J, Bast J, Clark J, et al.: Abundance, diversity, and distribution of mosquito vectors in selected ecological regions of Kenya: public health implications. J Vector Ecol. 2013; 38(1): 134–142. PubMed Abstract \| Publisher Full Text\n\nLutomiah JL, Koka H, Mutisya J, et al.: Ability of selected Kenyan mosquito (Diptera: Culicidae) species to transmit West Nile virus under laboratory conditions. J Med Entomol. 2011; 48(6): 1197–1201. PubMed Abstract \| Publisher Full Text\n\nMartens P, Hall L: Malaria on the move: human population movement and malaria transmission. Emerg Infect Dis. 2000; 6(2): 103–109. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMbogo CM, Mwangangi JM, Nzovu J, et al.: Spatial and temporal heterogeneity of Anopheles mosquitoes and Plasmodium falciparum transmission along the Kenyan coast. Am J Trop Med Hyg. 2003; 68(6): 734–742. PubMed Abstract \| Publisher Full Text\n\nMcDonald PT: Population characteristics of domestic Aedes aegypti (Diptera: culicidae) in villages on the Kenya Coast I. Adult survivorship and population size. J Med Entomol. 1977; 14(1): 42–48. PubMed Abstract \| Publisher Full Text\n\nMidega JT, Mbogo CM, Mwnambi H, et al.: Estimating dispersal and survival of Anopheles gambiae and Anopheles funestus along the Kenyan coast by using mark-release-recapture methods. J Med Entomol. 2007; 44(6): 923–929. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMidega JT, Muturi EJ, Baliraine FN, et al.: Population structure of Anopheles gambiae along the Kenyan coast. Acta Trop. 2010; 114(2): 103–108. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMixão V, Bravo Barriga D, Parreira R, et al.: Comparative morphological and molecular analysis confirms the presence of the West Nile virus mosquito vector, Culex univittatus, in the Iberian Peninsula. Parasit Vectors. 2016; 9(1): 601. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMonath TP: Dengue: the risk to developed and developing countries. Proc Natl Acad Sci U S A. 1994; 91(7): 2395–2400. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMorrill JC, Carpenter L, Taylor D, et al.: Further evaluation of a mutagen-attenuated Rift Valley fever vaccine in sheep. Vaccine. 1991; 9(1): 35–41. PubMed Abstract \| Publisher Full Text\n\nMurugan K, Vadivalagan C, Karthika P, et al.: DNA barcoding and molecular evolution of mosquito vectors of medical and veterinary importance. Parasitol Res. 2016; 115(1): 107–121. PubMed Abstract \| Publisher Full Text\n\nMuturi EJ, Shililu J, Jacob B, et al.: Mosquito species diversity and abundance in relation to land use in a riceland agroecosystem in Mwea, Kenya. J Vector Ecol. 2006; 31(1): 129–137. PubMed Abstract \| Publisher Full Text\n\nMwangangi JM, Mbogo CM, Orindi BO, et al.: Shifts in malaria vector species composition and transmission dynamics along the Kenyan coast over the past 20 years. Malar J. 2013; 12: 13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMwangangi JM, Midega J, Kahindi S, et al.: Mosquito species abundance and diversity in Malindi, Kenya and their potential implication in pathogen transmission. Parasitol Res. 2012; 110(1): 61–71. PubMed Abstract \| Publisher Full Text\n\nMwangangi J, Shililu J, Muturi E, et al.: Dynamics of immature stages of Anopheles arabiensis and other mosquito species (Diptera: Culicidae) in relation to rice cropping in a rice agro-ecosystem in Kenya. J Vector Ecol. 2006; 31(2): 245–251. PubMed Abstract \| Publisher Full Text\n\nNdiath MO, Mazenot C, Gaye A, et al.: Methods to collect Anopheles mosquitoes and evaluate malaria transmission: a comparative study in two villages in Senegal. Malar J. 2011; 10: 270. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNgugi HN, Mutuku FM, Ndenga BA, et al.: Characterization and productivity profiles of Aedes aegypti (L.) breeding habitats across rural and urban landscapes in western and coastal Kenya. Parasit Vectors. 2017; 10(1): 331. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPowell JR, Tabachnick WJ: History of domestication and spread of Aedes aegypti--a review. Mem Inst Oswaldo Cruz. 2013; 108(Suppl 1): 11–17. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nReisen WK, Meyer RP, Tempelis CH, et al.: Mosquito abundance and bionomics in residential communities in Orange and Los Angeles Counties, California. J Med Entomol. 1990; 27(3): 356–367. PubMed Abstract \| Publisher Full Text\n\nReisen WK, Reeves WC: Bionomics and ecology of Culex tarsalis and other potential mosquito vector species.1990; 254–329. Reference Source\n\nReiter P: Yellow fever and dengue: a threat to Europe? Euro Surveill. 2010; 15(10): 19509. PubMed Abstract\n\nRoberts D: Mosquitoes (Diptera:Culicidae) breeding in brackish water: female ovipositional preferences or larval survival? J Med Entomol. 1996; 33(4): 525–530. PubMed Abstract \| Publisher Full Text\n\nSabwa DM, Odindo MO, Otieno WA: Seasonal incidence of Amblyospora sp. (Thelohaniidae: Microsporidia) in Culex sitiens larvae at the Kenya coast. Int J Trop Insect Sci. 1984; 5(04): 269–272. Publisher Full Text\n\nScarpassa VM, Cunha-Machado AS, Saraiva JF: Evidence of new species for malaria vector Anopheles nuneztovari sensu lato in the Brazilian Amazon region. Malar J. 2016; 15: 205. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchäfer M, Lundström JO: Comparison of mosquito (diptera: culicidae) fauna characteristics of forested wetlands in sweden. Acta Neophilologica. 2001; 94(4): 576–582. Publisher Full Text\n\nSemenza JC: Climate change and human health. Int J Environ Res Public Health. 2014; 11(7): 7347–7353. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTeesdale C: Studies on the Bionomics of Aëdes aegypti (L.) in its Natural Habitats in a Coastal Region of Kenya. Bus Econ Rev. 1955; 46(03): 711–742. Publisher Full Text\n\nThenmozhi V, Rajendran R, Ayanar K, et al.: Long-term study of Japanese encephalitis virus infection in Anopheles subpictus in Cuddalore district, Tamil Nadu, South India. Trop Med Int Health. 2006; 11(3): 288–293. PubMed Abstract \| Publisher Full Text\n\nThompson JD, Gibson TJ, Plewniak F, et al.: The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 1997; 25(24): 4876–4882. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTurell MJ, O’Guinn M, Oliver J: Potential for New York mosquitoes to transmit West Nile virus. Am J Trop Med Hyg. 2000; 62(3): 413–414. PubMed Abstract \| Publisher Full Text\n\nVanlandingham DL, Hong C, Klingler K, et al.: Differential infectivities of o'nyong-nyong and chikungunya virus isolates in Anopheles gambiae and Aedes aegypti mosquitoes. Am J Trop Med Hyg. 2005; 72(5): 616–621. PubMed Abstract \| Publisher Full Text\n\nWaterhouse AM, Procter JB, Martin DM, et al.: Jalview Version 2--a multiple sequence alignment editor and analysis workbench. Bioinformatics. 2009; 25(9): 1189–1191. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWesolowski A, Eagle N, Tatem AJ, et al.: Quantifying the impact of human mobility on malaria. Science. 2012; 338(6104): 267–270. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWHO: In Kenya, the path to elimination of malaria is lined with good preventions. Retrieved February 4, 2019. 2017. Reference Source\n\nWHO: World malaria report 2017. WHO. 2018. Reference Source\n\nWilder-Smith A, Gubler DJ: Geographic expansion of dengue: the impact of international travel. Med Clin North Am. 2008; 92(6): 1377–90, x. PubMed Abstract \| Publisher Full Text\n\nWoolhouse ME, Dye C, Etard JF, et al.: Heterogeneities in the transmission of infectious agents: implications for the design of control programs. Proc Natl Acad Sci U S A. 1997; 94(1): 338–342. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "47920", "date": "13 May 2019", "name": "Nicholas Johnson", "expertise": [ "Reviewer Expertise Zoonotic viruses" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article is a focused investigation of the haplotype variation observed in mosquito populations trapped at three locations in Kenya. The authors noted variation in all the 11 species reported including a number of novel observations. However, the authors do not include basic data on the actual distribution and species assemblage at the collection sites. By arbitrarily selecting 25 samples for extensive genetic analysis and ignoring the remaining samples that apparently included 894, 824 and 720 mosquitoes appears to be a fundamental omission. It is difficult to see how the authors can conclude that “The distribution varies in density” when the dataset has not been analysed. Whilst it may be beyond the resources of the team to molecularly type all 2,438 samples, without some attempt to include morphological identification of a significant proportion of these samples the manuscript is considerable diminished.\nThe authors should check the capitalisation of pathogens throughout. As a general rule, names derived from a place are capitalised e.g. Rift Valley fever virus, whilst those that are not are in lower case e.g. yellow fever virus, malaria.\nThe authors must revise the conclusions section to reflect the findings of the paper stating precisely what they have derived from their observations. At the moment the two sentences’ provide a revision of the manuscripts aim and a vague statement that is unsupported by the results.\nThe reference for Morrill et al is incorrect. It should be Morrill et al., 1991, J Trop Med Hyg, 94, 166.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4915", "date": "24 Sep 2019", "name": "Moni Makanda", "role": "Author Response", "response": "First and foremost thank you for taking time to review the paper, your views were most welcomed and addressed. It was beyond our financial capability to genetically analyse 2,438 samples. Earlier study by (Hale, 2012) support my study as adequate in phylogenetic analysis. However, morphological identification was done and further analysis by use of PCR-HRM. This data has been capture in my MSc. thesis, moreover a second publication on the same is underway. As this paper was focused on genetic diversity we focused on molecular analysis. Names of pathogens throughout the article have been corrected based on the general rule. The conclusion was strengthened as proposed. Reference Morril et al. 1991 was revised as advised" } ] }, { "id": "47313", "date": "28 May 2019", "name": "Laban Njoroge", "expertise": [ "Reviewer Expertise I am a mosquito taxonomist and ecologist. I am not an expert in sequencing and phylogenetic analyses and therefore that bit may require another expert." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAbstract and Introduction: There are a number of grammatical errors in the manuscript. The following are examples in the abstract: 'In this study, we identified mosquito species across Kisumu, Kilifi and Nairobi Counties' - I believe mosquitoes were collected from only few places in the three counties and therefore should read were collected from and not across. There should be a word 'the' before consequent mortality in line 1 of the abstract. Commas are missing in some areas where they should be e.g. in the abstract. '.........Despite reports on increase of new and recurrent mosquito borne-disease outbreaks such as chikungunya, dengue fever and Rift valley fever in Kenya little is known about the genetic characteristics.' A comma after Kenya.\n'PCR was used to amplify and sequence the partial cytochrome oxidase subunit 1 (CO1) gene.' This sentence creates an impression that PCR was used for both amplificationa nd sequencing. I believe PCR was used to amplify the genes but sequencing was done using another method. This should be clarified. It is needless to repeat the word haplotypes three types. In the introduction: '.............pathogens causing diseases such as; malaria, lymphatic filariases, avian malaria' - That semi colon after such as should not be there. There is lack of standardization in writing names. e.g. disease names such as Chikungunya, Dengue are started with a capital letter in the main body but with small letters in the Keywords Some scientific names of some species are wrongly written e.g. Aedes cummnisii should be Aedes cumminsii. Culicine mosquitoes not Culcine mosquitoes in some areas and others mosquitos.\nMethods: There aren't sufficient details of the methods to allow replication by others. For instance, there is no mention of where the Pyrethrum Spray Catches were done. Was in inside houses, Bus waiting lounges, garages? The reference for PSC collection method should clearly indicate that it is as used by Ndiath et al. 2011. I have rated the study design study as partly appropriate as being a study that targeted diversity, one sampling method that targets only indoor resting mosquitoes was not the best. There is probably a need to justify why only PSC was used and point one method as the reason for the low diversity collected. Again, it is not explained why bus stops in the three counties were preferred as sampling sites. If the idea was to see the contribution of transportation to the mixing of populations, then that didn't come out clearly.\n\n'Only a few Aedes aegypti in Nairobi (Kinuthia et al., 2017)'. The authors do not tell us it is few of what. Few haplotypes or individuals? diversity and spread of Aedes aegypti has been associated with expansion on arboviral infection...........' - This statement needs to be rephrased. The word on can be replaced by of. Alternatively, it can be Diversity and spread of Ae. aegypti has been attributed to the increase in arboviral infections.\n'and was predominated in the Kilifi samples......' Should read it was predominant in Kilifi samples.\n'This may contribute to the high susceptibility to dengue-outbreak reported in the region (Baba et al' - Should read this may contribute to their high susceptibility 'The similarities in the genetic composition between the An. gambiae in Kenya and Uganda is most likely due to the proximity of the two countries to one another and the exchanges is more likely over land as opposed to across lake Victoria as claimed in the discussion. This study has indicated high diversity of Anopheles haplotypes in the Kisumu population' - I do not think two species only can be regarded as high diversity. Probably you should use the word higher in comparison with Nairobi and Kilifi. 'The low diversity of Anopheles species in Kilifi and Nairobi may be attributed to the Great Rift Valley,' - there is an abundance of Anopheles especially in Kilifi (see Mwangangi et al 2012 which you have in the references). The problem is the choice of sampling method employed. PSC targets indoor resting mosquitoes only while the highest diversity are found outdoors\nConclusion: The conclusion is sounding a bit weak and it is more of a discussion than a conclusion. There isn't a strong conclusion about the findings on diversity and molecular characterization of mosquitoes encountered.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4916", "date": "24 Sep 2019", "name": "Moni Makanda", "role": "Author Response", "response": "First and foremost I would like to thank you for taking your time to review this article. Your views were most welcome and addressed accordingly. Grammatical errors were corrected throughout the article as highlighted. Effect of transportation on mosquito diversity was not a focus in this study. This study looked in to the diversity of mosquitoes in the study sites being town areas. The conclusion was strengthened as recommended." } ] } ]	1	https://f1000research.com/articles/8-262
https://f1000research.com/articles/7-1197/v1	06 Aug 18	{ "type": "Research Article", "title": "NLRX1 is not involved in the host defense against Escherichia coli induced pyelonephritis", "authors": [ "Lotte Kors", "Loes M. Butter", "Nike Claessen", "Gwendoline J.D. Teske", "Stephen E. Girardin", "Sandrine Florquin", "Jaklien C. Leemans", "Loes M. Butter", "Nike Claessen", "Gwendoline J.D. Teske", "Stephen E. Girardin", "Sandrine Florquin", "Jaklien C. Leemans" ], "abstract": "Background: Urinary tract infections (UTIs) caused by uropathogenic Escherichia coli (E. coli) are one of the most prominent infections that have serious impact on kidney functioning and the development of chronic kidney disease. NOD-like receptor (NLR)X1 is an innate immune receptor that is important for immune metabolism and regulation, with as yet an unknown role in UTI and the pathophysiology of pyelonephritis. Methods: Wild-type (WT) and NLRX1 Knock-out (KO) female mice were subjected to UTI by intravesically inoculation of uropathogenic E. coli and sacrificed at 24h and 48h after infection after which bacterial burden and the inflammatory response in the bladder and kidney were studied. Ex vivo we studied the role of NLRX1 during the LPS induced pro-inflammatory cytokine response and phagocytosis of E. coli by granulocytes and monocytes. Results: Here, we report that during early experimental UTI NLRX1 absence reduces bacterial clearance in the bladder and dampens the inflammatory cytokine response, whereas in the kidney NLRX1 does not affect bacterial burden or cytokine response. In addition, we found that NLRX1 is not essential for the pro-inflammatory cytokine secretion by granulocytes and monocytes in response to LPS nor for bacterial phagocytosis. Conclusion: Together, we report that NLRX1 is important in enhancing the early host defense against uropathogenic E. coli in the bladder but does not affect the development of pyelonephritis.", "keywords": [ "Innate Immune Receptor NLRX1", "Lower and Upper UTI", "Pyelonephritis", "Animal model", "Escherichia coli" ], "content": "Introduction\n\nToll like receptors (TLRs) and NOD-like receptors (NLRs) are members of a large family of extracellular and intracellular pattern recognition receptors (PRRs) that trigger immune responses to prevent pathogen invasion and growth1,2. Urinary tract infections (UTIs) are common bacterial infections in humans, that occur most commonly in women and children3. UTIs are caused by the presence of uropathogenic bacteria, usually Escherichia coli (E. coli), in the lower urinary tract (bladder) that overcome the host innate immune defense. When the infection ascends from the bladder via the ureters to the upper renal pyelum, lower UTI can lead to acute pyelonephritis. If untreated pyelonephritis can have serious implications for renal functioning and the development of damage and scarring4,5. Antimicrobial resistance among UTIs are increasing6,7, therefore new insights in host defense mechanisms are required to obtain new targets for therapy.\n\nTLRs are known to play an important role in the host response to UTIs8, whereas the role of NLRs herein is unclear. NOD-like receptor X1 (NLRX1) is an ubiquitously expressed PRR in mitochondria that controls mitochondrial activity in tubular epithelial cells and hepatocytes, and in this way effects respectively ischemic acute kidney disease and liver steatosis9,10. Other functions for NLRX1 include negative regulation of antiviral immunity11, and inhibition of NF-κB signaling by disrupting interaction of TRAF6 and IKK12,13. Given these studies, NLRX1 could play a potential role during the pathophysiology of acute bacterial infections such as pyelonephritis.\n\nTo get more insight in NLRX1 functioning during bacterial infection we investigated in the present study the role of NLRX1 during uropathogenic E. coli-induced lower and upper UTI in mice. We found that although NLRX1 absence enhances bacterial burden in the bladder during the early phase of infection, NLRX1 is not involved in the host defense against pyelonephritis.\n\n\nMethods\n\nNLRX1 KO mice with a C57BL6/J background were generated as described previously14 and bred at the animal facility of the Academic Medical Center (AMC) in Amsterdam, The Netherlands. Age- and gender-matched C57BL6/J WT mice were obtained from Charles River (Maastricht, The Netherlands). Animals were housed in individual ventilated cages (IVCs) with bedding and cage enrichment that were kept under standard environmental conditions (temperature, humidity, ventilation, light/dark cycle) and under specific pathogen-free conditions (SPF) with ad libitum access to water and food. The mice were allowed to acclimatize for a week before starting the experimental procedures.\n\nThe in vivo study was performed with 2 experimental groups: 1) WT (n=8) and 2) NLRX1-KO (n=8) and 2 sham/control groups: 3) WT (n=4) 4) NLRX1-KO (n=4). Each experimental group was subjected for two time points (24h and 48h) to UTI as described previously15 and briefly explained later. The total number of mice per in vivo experimental group was 16 and the total number per sham/control group was 4. To reach a statistical significant effect of NLRX1 deficiency the number of 8 mice per experimental group was based on a variation coefficient of 15%, a minimal relative effect of 30%, a P value of 5% and a power of 80%, that were based on previous studies done in our group15–17. For both experimental and sham/control groups 11–12 week old female mice (median weights: WT; 19,6 and NLRX1-KO; 21,3 grams) were used. Each experimental group was divided in 2 cages of 3 and 5 animals and in the sham/control group 4 animals per cage were kept. For the experimental group uropathogenic E. coli 1677, isolated from an uroseptic patient, was cultured in the laboratory in sterile Tryptic Soy Broth (TSB) overnight at 37°C, 5% CO2. The next day, in the morning this suspension was diluted 1:100 in fresh TSB and in 2–3h cultured to optical density OD620nm = 1 was reached (measured with a spectrophotometer (DU640, Beckman, USA)). Subsequently bacteria were spun down for 14 min at 4000 rpm at 4°C, washed three times and resuspended in 10 mL sterile PBS. The same day, in the animal facility, under general anaesthesia (10 µl/1 g body weight of FFM mixture, containing 1.25 mg/ml midazolam (Dormicum®, Roche, Woerden, The Netherlands), 0.08 mg/ml fentanyl citrate/2.5 mg/ml fluanisone (Hypnorm, Veta Pharma Ltd., Leeds, UK)) that was given intraperitoneal, mice were via the urethra intravesically inoculated with 8108 CFU in a 100µl volume. Mice in the sham/control group underwent the same procedure with administration of 100µl sterile PBS. CFU concentrations in the inoculum were determined by plating 10-fold serial dilutions on blood-agar plates at 37°C, 5% CO2 overnight. Mice were sacrificed 24 and 48 hour post-inoculation by cardiac puncture under 4% isoflurane/O2 followed by cervical dislocation. Blood was collected in lithium-heparin containing tubes and kidneys and bladders were collected for further analysis. In the WT 24h group one animal reached, because of signs of severe sepsis, the human end point and was excluded from further data analysis. The animals used to study leukocyte composition and ex vivo granulocyte and monocyte functioning contained 2 experimental groups: 1) WT (n=6) and 2) NLRX1-KO (n=6) and were sacrificed as described earlier for the in vivo experiments. Mice used for the ex vivo experimental groups were 13–14 week old female mice (median weights: WT;26.5 and NLRX1-KO;26.6 grams). The total number of mice per ex vivo experimental group was 6. In the WT group one sample was excluded from further data analysis due to low blood gain.\n\nAll animal procedures were ethically approved under DPA 25 AB-1 by the Animal Care and Use Committee of the Academic Medical Center Amsterdam and were conducted in compliance with the ARRIVE guidelines (NC3Rs).\n\nBladder 20% (w/v) and left kidney 10% (w/v) tissues were homogenized in sterile PBS with a tissue homogenizer (Polytron PT1300D homogenizer, Kinematica AG). To determine bacterial loads 10-fold serial dilutions of bladder- and kidney-homogenates were plated onto blood agar plates. Colonies were counted 16h after incubation at 37°C.\n\nAbsolute leukocyte number in blood was measured with a Coulter Counter (Beckmann Coulter Inc., Fullerton, CA). To assess relative leukocyte composition, 100 µl whole blood erythrocytes were lysed by adding 2 ml lysis buffer (8.3g/L NH4Cl, 1.0g/L KHCO3, 0.1 mM EDTA, pH 7.4) for 15 min at RT. The remaining cells, leukocytes, were washed once, centrifuged and resuspended in FACS buffer (0.5% BSA, 0.01% NaN3, 0.35mM EDTA in PBS) and were measured by flow cytometry on a FACS Canto (BD Biosciences, San Jose, CA, USA) and analyzed with FlowJo version 10 software. Living cells, lymphocyte, granulocyte and monocyte populations were gated based on forward-scattered light (FSC)/side-scattered light (SCC).\n\nWhole blood of WT and NLRX1 KO mice was incubated at 37oC, 5% CO2 in 10ng/mL LPS (cat no. l4391, Sigma-Aldrich, Zwijndrecht, The Netherlands) conditioned RPMI medium (Thermo fisher Scientific, Waltham, MA, USA) containing 10% FCS (Invitrogen, Carlsbad, CA, USA) with 2 mM L-glutamine and 100 U/mL penicillin/streptomycin (all from Thermo fisher Scientific, Waltham, MA, USA). After 14h, cells were spun down (5 min, 4000 rpm) and supernatants were collected and stored at -20°C prior to use for cytokine measurements.\n\nPhagocytosis by granulocytes and monocytes was measured with the PHAGOTEST (cat no. 10–0100, Glycotope Biotechnology, Heidelberg, Germany) according to the manufacturer’s instructions. Briefly, 100 µl of heparinized whole blood was incubated with opsonized FITC- labeled E. coli for 10 minutes in a 37°C water bath, whereas the negative controls remained on ice. After phagocytosis was stopped the surface signal was quenched. After lysing the erythrocytes and fixation of the leukocytes, phagocytic capacity was measured by flow cytometry on a FACS Canto (BD Biosciences, San Jose, CA, USA) and analyzed with FlowJo version 10 software.\n\nBladder 5% (w/v) and kidney 10% (w/v) tissues from 24h and 48h UTI subjected WT and NLRX1 KO mice were homogenized with a tissue homogenizer (Polytron PT1300D homogenizer, Kinematica AG) in Greenberger lysis buffer (GLB) (300mM NaCl, 30mM Tris, 2mM MgCl2, 2mM CaCl2, 1% (v/v) Triton X-100, pH set at 7.4, supplemented with Protease Inhibitor Cocktail II (cat no. p8340, Sigma-Aldrich, Sigma-Aldrich, Zwijndrecht, The Netherlands). Levels of keratinocyte-derived chemokine (KC), macrophage inflammatory protein-2 (MIP-2), monocyte chemoattractant protein 1 (MCP-1), interleukin 1β (IL1β), interleukin 6 (IL6), tumor necrosis factor alpha (TNFα) and mouse myeloid peroxidase (MPO) were determined in bladder and kidney homogenates and whole blood plasma (KC, TNFα and IL6 only) by duo set ELISAs (cat no. MKC00B, MM200, MJE00, MLB00C, M6000B, MTA00B, DY3667, R&D Systems, Abingdon, UK), performed according to the supplied protocols. ELISA data measured in bladder and kidney homogenates was adjusted for total protein concentration as determined by BCA protein assay (cat no. B9643, Sigma-Aldrich, Sigma-Aldrich, Zwijndrecht, The Netherlands) developed with 4% CuSO4.\n\nFrom hepanarized blood, plasma was obtained from the upper phase after spinning the tube for 10 minutes at 2000 rpm. Urea and creatinine levels in plasma were determined at room temperature by colorimetric enzyme reactions involving creatinase and urease and analyzed on the Cobas c702 Chemistry Analyzer (Roche Diagnostics, Indianapolis, IN, USA) according to standard diagnostic procedures performed by the department of Clinical Chemistry of the Academic Medical Center Amsterdam.\n\nTotal RNA was extracted from snap-frozen -80°C stored bladder and kidney using TriReagent (cat no. T9424, Sigma-Aldrich, Zwijndrecht, The Netherlands) followed by chloroform phase separation to obtain the aqueous RNA containing upper phase and isopropanol precipitation according to the manufacturer protocol procedure description and converted to cDNA. cDNA was synthesized using M-MLV reverse transcriptase according the procedure described in the manufacturer protocol (cat no. 28025, Thermo Scientific). Transcription was analyzed by RT-qPCR on a Roche LightCycler 480 using 2.5 µl sensiFAST SYBR master mix (cat no. bio-98020, Bioline reagents, London, UK), 0.20 µl forward primer, 0.20 µl reverse primer (Table 1), 2.10 µl distilled H2O and 1µl cDNA per reaction. qPCR primers were synthesized by Eurogentec (Maastricht, The Netherlands) and described in the list below. qPCR data was analyzed based on linear regression using the LinRegPCR program, that is freely available18,19. Briefly, the LinRegPCR program imports non-baseline corrected qPCR data, performs a baseline correction on each sample separately, determines a window-of-linearity and then uses linear regression analysis to determine the PCR efficiency per sample from the slope of the regression line. The mean PCR efficiency per amplicon and the Cq value per sample are used to calculate a starting concentration per sample, expressed in arbitrary fluorescence units (au)19.\n\nData are expressed as mean ± standard error of the mean (SEM), bacterial outgrowth data are expressed on a logarithmic scale as median scatterplot. The non-parametric Mann Whitney U test was performed for two group comparison. For all analyses, values of P ≤ 0.05 were considered significant. All statistical analyses were performed using GraphPad Prism 5 (GraphPad Software, San Diego, CA, USA).\n\n\nResults\n\nTo determine whether NLRX1 expression is modulated in the murine bladder and kidney during urinary tract infection (UTI), wild-type (WT) mice were intravesically inoculated with uropathogenic E. coli and sacrificed at 24h and 48h after infection. Non-infected sham mice were used as controls. Real-time quantitative PCR revealed that Nlrx1 transcript levels were constitutively present in the bladder and kidney (Figure 1A and B). Nlrx1 transcript levels show a non-significant trend towards increased levels in the bladder after 24h, while levels are returned towards baseline sham levels at 48h (Figure 1A). In the kidney Nlrx1 transcript levels remained at baseline level 24h after infection while after 48h levels were significantly increased (Figure 1B). Together, these data show that in response to UTI, local Nlrx1 expression is increased upon E. coli infection.\n\nNlrx1 mRNA transcript levels in wild-type (WT) (A) bladder and (B) kidney from sham, and uropathogenic E. coli-inoculated mice. All data are expressed as mean ± SEM, n=4 (sham) and n=7-8 (24h and 48h) animals per group. Statistical significance was determined by non-parametric Mann Whitney U test, P<0.05.\n\nTo investigate whether NLRX1 plays a role in bladder and kidney during lower and upper UTI, we examined bacterial loads in these organs from WT and NLRX1 knock-out (KO) mice 24h and 48 h after inoculation with uropathogenic E. coli. This revealed that the bacterial outgrowth, as measured by colony forming units (CFU), in bladder tissue from NLRX1 KO mice was significantly higher at 24h after infection compared to WT while at 48h no differences in bacterial burden were found (Figure 2A). NLRX1 deficient mice had more improved bacterial clearance from the bladder at 48h as compared to 24h (Figure 2A). No differences in the amount of CFU were found between kidneys from WT and NLRX1 KO mice at both time points (Figure 2B). To monitor the local inflammatory response during infection, we next measured the levels of KC, MIP-2, MCP-1, IL-1β, IL-6 and TNFα in kidney and bladder homogenates (Figure 2C and D). The production of MIP-2 in the bladder was in NLRX1 KO mice reduced at both 24h and 48h after infection compared to WT animals (Figure 2C). In addition, NLRX1 KO bladders show reduced levels of MCP-1 and TNFα compared to WT at 48h, while no differences were found in KC, IL-1β and IL6 levels (Figure 2C). We identified an increase in renal KC levels at 24h in NLRX1 KO mice compared WT mice, whereas at 48h KC levels were similar between both groups (Figure 2D). At both 24h and 48h no differences were found in renal MIP-2, MCP-1, IL-1β, IL-6 and TNFα levels between WT and NLRX1 KO mice (Figure 2D). In addition, upper UTI and NLRX1 have no significant influence on renal function as reflected by similar plasma levels of urea and creatinine between all mice (Supplementary Figure 1A and B). Together, this indicates that the lack of NLRX1 is associated with an early impaired ability to clear uropathogenic E. coli from the bladder, probably due to an impaired pro-inflammatory cytokine response, while NLRX1 deficient mice had an improved bacterial clearance from 24h to 48h in the bladder as compared to WT. No differences were found in bacterial burden and cytokine response in the kidney when mice were deficient for NLRX1.\n\nOutgrowth of uropathogenic E. coli expressed in colony forming units (CFU) in (A) bladder and (B) kidney homogenates from wild-type (WT) and NLRX1 knock-out (KO) mice 24h and 48h after inoculation. Levels of KC, MIP-2, MCP-1, IL-1β, IL6 and TNFα in (C) bladder and (D) kidney homogenates from the E. coli inoculated WT (white bars) and NLRX1 KO (black bars) mice. Data at A and B are expressed on a logarithmic scale as median scatterplot. Data at C and D are expressed as mean ± SEM. For all data n=7-8 animals per group and statistical significance between WT and NLRX1 KO was determined by non-parametric Mann Whitney U test, P<0.05 and P<0.01.\n\nBy analyzing inflammatory cells in the circulation we observed that the numbers of granulocytes and monocytes were equal between uninfected WT and NLRX1 KO mice (Figure 3A). A non-significant trend towards an increased presence of lymphocytes in NLRX1 KO compared to WT is shown (Figure 3A). We observed that WT and NLRX1 KO mice 24h and 48h after inoculation, have similar numbers of circulating leukocytes (Figure 3B). Recruitment of neutrophils in the kidney and bladder are essential for the host defense against uropathogenic E. coli20. Therefore, we determined active neutrophil presence by measuring kidney and bladder myeloperoxidase (MPO) concentrations. No differences in MPO levels were found in bladder and kidney between WT and NLRX1 KO mice at 24h and 48h (Figure 3C and D), indicating a similar number of activated neutrophils.\n\n(A) Total circulating leukocyte, lymphocyte, granulocyte and monocyte cell counts in blood from uninfected wild-type (WT) (white bars) and NLRX1 knock-out (KO) (black bars) mice n=5–6 animals per group. (B) Circulating leukocytes in blood from WT (white bars) and NLRX1 KO (black bars) mice 24h and 48h after uropathogenic E. coli inoculation. MPO levels in (C) bladder and (D) kidney homogenates from 24h and 48h inoculated WT (white bars) and NLRX1 KO mice (black bars). All data are expressed as mean ± SEM, n=7–8 (B–D) animals per group. Statistical significance between WT and NLRX1 KO was determined by non-parametric Mann Whitney U test.\n\nSince we observed differences in bacterial outgrowth in the bladder while the number of local neutrophils after infection is equal between WT and NLRX1 KO, we investigated if NLRX1 absence causes functional changes to granulocytes and monocytes. This revealed that NLRX1 is not critical for the secretion of the pro-inflammatory cytokines KC, TNFα and IL6 after ex vivo whole blood LPS stimulation (Figure 4A, B and C). To investigate if NLRX1 is important for the phagocytic activity of granulocytes and monocytes, leukocytes from WT and NLRX1 KO mice were ex vivo incubated with fluorescein labelled opsonized E. coli, and phagocytosis was analyzed using flow cytometry. Granulocytes and monocytes from both WT and NLRX1 KO mice show increased phagocytic activity responses when challenged with E. coli at 37°C compared to 0°C (Figure 4D and E). However, no differences were observed in the percentage of granulocytes and monocytes that undergo phagocytosis between WT and NLRX1 KO (Figure 4E and F). Together, these results suggest that the early decreased bacterial clearance in the bladders from NLRX1 KO mice cannot be explained by an impaired granulocyte or monocyte response.\n\nLevels of (A) KC (B) TNFα and (C) IL6 after 14h ex vivo LPS whole blood stimulation from wild-type (WT) (white bars) and NLRX1 knock-out (KO) (black bars). ND = not detectable. Phagocytic activity responses of (D) granulocytes and (E) monocytes from WT (white bars) and NLRX1 KO (black bars) mice ex vivo challenged with E. coli at 0°C and 37°C. All data are expressed as mean ± SEM, n=5–6 animals per group. Statistical significance between all columns was determined by non-parametric Mann Whitney U test, P<0.05 and **P<0.01.\n\n\nDiscussion\n\nInnate immune receptors like TLRs and NLRs are known to play pivotal roles in the first line of host defense against invading pathogens. NLRX1 is an innate immune receptor that can modulate inflammatory responses21 and cell metabolism9,10. As such NLRX1 could play a potential role during the pathophysiology of UTI. To study this we investigated the role of NLRX1 during uropathogenic E. coli-induced lower and upper UTI in mice. Although NLRX1 enhances the inflammatory cytokine response and the bacterial clearance in the bladder during early experimental UTI, we found that this receptor does not affect overall renal bacterial loads and inflammation during pyelonephritis. In addition, we observed that NLRX1 is not essential for pro-inflammatory cytokine secretion by granulocytes and monocytes in response to LPS nor for phagocytosis of E. coli.\n\nIn this study we investigated the role of NLRX1 in influencing bacterial burden and inflammation in bladder and kidney during experimental UTI. We found that the lack of NLRX1 is associated with an impaired ability to clear uropathogenic E. coli from the bladder at 24h only, probably due to an impaired pro-inflammatory MIP-2 cytokine response which is usually needed for the recruitment of granulocytes to the site of infection20,22 and the initiation of host defense during UTI23. Surprisingly however, the levels of bladder MPO as an indicator of neutrophil influx and the ex vivo granulocyte phagocytic capacity to ingest E. coli are not affected by NLRX1 while local Nlrx1 expression in the bladder tended to be increased at 24h. Possibly, granulocyte influx is altered by NLRX1 at an earlier time point than 24h. Despite the impaired MIP-2, MCP-1 and TNFα cytokine response, bacterial burden in the bladder at 48h, is not different in NLRX1 deficient mice compared to WT. In fact, NLRX1 KO are able to clear E. coli bacteria faster since the CFU reduction from 24h to 48h is significantly while bacterial levels in bladder remained the same in WT. Whereas in the kidney, the outgrowth from 24h to 48h in WT and NLRX1 KO remains unchanged, indicating that despite the local Nlrx1 expression increase at 48h NLRX1 this does not affect bacterial outgrowth in the kidney. Our study demonstrates that due to NLRX1 absence, the MIP-2 cytokine release to recruit neutrophils is less pronounced and hence possibly attenuates the early phase of the host defense against E. coli in the bladder without affecting later bacterial bladder burden, innate myeloid cell phagocytosis and the promotion of pyelonephritis.\n\nNLRX1 is on the one hand described to negatively regulate NF-κB signaling12,13 and on the other to indirectly amplify the NF-κB pathway24. During E. coli-induced UTI infections the NF-κB signaling pathway is via TLR4 activation one of the most important pathways for the pro-inflammatory cytokine release and the clearance of E. coli from the urinary tract25. We observed in NLRX1 KO bladders that despite the increased bacterial burden 24h post infection, besides reduced MIP-2, the levels of pro-inflammatory cytokines TNFα, IL6 and IL1β were equal in both mouse strains. Whereas at the later 48h time point in the bladder and at 24h and 48h in the kidney, NLRX1 has no effect on pro-inflammatory cytokine response and bacterial burden. This indicates that upon early UTI, NLRX1 absence leads to a suppressed pro-inflammatory cytokine response in the bladder. Whether this is caused by an altered neutrophil influx in the early onset of infection or the ability of NLRX1 to influence NF-κB signaling warrants further study. From our study it is however clear that NLRX1 is neither essential for whole blood pro-inflammatory KC, TNFα and IL6 cytokine responses to LPS, nor for bacterial phagocytosis by granulocytes and monocytes. Similar observations were done in bone marrow-derived macrophages where TNFα and IL6 cytokine expression remained similar in WT and NLRX1 deficient cells after a Helicobacter pylori (LPS positive) infection26. In contrast, TNFα and IL6 levels were increased upon NLRX1 knockdown in LPS-stimulated peritoneal macrophages12 and IL6 levels in LPS-stimulated bone marrow-derived macrophages13 indicating that NLRX1 attenuates NF-κB signaling. In contrast, increased NF-κB signaling upon LPS positive Shigella flexineri infection was observed in NLRX1 overexpressing epithelial cells24. Possibly, the role of NLRX1 varies between cell types involved in host defense, such as myeloid cells and parenchymal cells, during different time points of infection and different ligands. Together, we observed that NLRX1 does not affect the pro-inflammatory cytokine response after LPS challenge in granulocytes and monocytes, whereas previous studies show that in macrophages and epithelial cells NLRX1 can behave differently12,13,24. Indeed, during UTI epithelial cells are important in activating inflammation via various signaling pathways25. Based on our results that granulocyte and macrophage functioning are not affected by NLRX1, we assume that during UTI NLRX1 plays a role in the early bacterial burden in the bladder by activating the pro-inflammatory cytokine response in parenchymal cells via NF-κB.\n\nBesides its role in immune regulation, we previously observed that NLRX1 functioning extends to the control of mitochondrial activity, oxidative stress and cellular metabolism in parenchymal cells of the kidney and liver9,10. Macrophages and in particular neutrophils contribute during infections to the host defense via oxidative burst27,28. From our data it is not clear if NLRX1 plays a role in the oxidative burst in myeloid cells during UTI. However, a previous study showed that NLRX1 plays no significant role in ROS production of LPS activated neutrophils and macrophages13. Together, our data indicates that there is a role for NLRX1 during UTI in the bladder by activating the pro-inflammatory cytokine response, while no direct role for NLRX1 is observed in myeloid cells.\n\n\nConclusion\n\nWe report that NLRX1 is important in attenuating the early bacterial burden in the bladder by enhancing the local pro-inflammatory cytokine response but has no effect on the development of pyelonephritis.\n\n\nData availability\n\nDataset 1: NLRX1 expression data – This file contains the data underlying the analysis of the NLRX1 expression in bladder and kidney as shown in Figure 1. 10.5256/f1000research.15361.d21293729\n\nDataset 2: In vivo mice bladder and kidney colony forming units (CFU)-, cytokine-, general marker-data and leukocyte counts in sham and after 24h and 48h of infection – This file contains the data underlying the analysis of the data used in Figure 2, Figure 3 and Supplementary Figure 1. 10.5256/f1000research.15361.d21293830\n\nDataset 3: Ex vivo FACS output data - This file contains the FACS data underlying the leukocyte composition analysis as shown in Figure 3. 10.5256/f1000research.15361.d21293931\n\nDataset 4: Ex vivo cytokine data – This file contains the data underlying the cytokine determination in LPS stimulated whole blood as shown in Figure 4. 10.5256/f1000research.15361.d21294032\n\nDataset 5: Ex vivo FACS output data on granuocyte and monocytes phagocytosis - This file contains the FACS data underlying the granuocytes and monocytes phagocytosis as shown in Figure 4. 10.5256/f1000research.15361.d21294133", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis study was supported by The Netherlands Organization for Scientific Research NWO grant [91712386] to JCL.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nSupplementary Figure 1. Renal function of wild-type (WT) and NLRX1 knock-out (KO) mice during experimental urinary tract infections (UTI)\n\nPlasma levels of renal function markers (A) urea and (B) creatinine from sham, 24h and 48h uropathogenic Escherichia coli inoculated WT (white bars) and NLRX1 KO (black bars) mice. All data are expressed as mean ± SEM, n=3–4 (sham) and n=7–8 (24h and 48h) animals per group. Statistical significance was determined by non-parametric Mann Whitney U test.\n\nClick here to access the data.\n\n\nReferences\n\nKawai T, Akira S: The roles of TLRs, RLRs and NLRs in pathogen recognition. Int Immunol. 2009; 21(4): 317–337. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCreagh EM, O’Neill LA: TLRs, NLRs and RLRs: a trinity of pathogen sensors that co-operate in innate immunity. Trends Immunol. 2006; 27(8): 352–357. PubMed Abstract \| Publisher Full Text\n\nFoxman B: The epidemiology of urinary tract infection. Nat Rev Urol. 2010; 7(12): 653–660. PubMed Abstract \| Publisher Full Text\n\nPellé G, Vimont S, Levy PP, et al.: Acute pyelonephritis represents a risk factor impairing long-term kidney graft function. Am J Transplant. 2007; 7(4): 899–907. PubMed Abstract \| Publisher Full Text\n\nRagnarsdottir B, Lutay N, Grönberg-Hernandez J, et al.: Genetics of innate immunity and UTI susceptibility. Nat Rev Urol. 2011; 8(8): 449–468. PubMed Abstract \| Publisher Full Text\n\nShepherd AK, Pottinger PS: Management of Urinary Tract Infections in the Era of Increasing Antimicrobial Resistance. Med Clin North Am. 2013; 97(4): 737–757, xii. PubMed Abstract \| Publisher Full Text\n\nSanchez GV, Master RN, Karlowsky JA, et al.: In vitro antimicrobial resistance of urinary Escherichia coli isolates among U.S. outpatients from 2000 to 2010. Antimicrob Agents Chemother. 2012; 56(4): 2181–2183. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeemans JC, Kors L, Anders HJ, et al.: Pattern recognition receptors and the inflammasome in kidney disease. Nat Rev Nephrol. 2014; 10(7): 398–414. PubMed Abstract \| Publisher Full Text\n\nKors L, Rampanelli E, Stokman G, et al.: Deletion of NLRX1 increases fatty acid metabolism and prevents diet-induced hepatic steatosis and metabolic syndrome. Biochim Biophys Acta. 2018; 1864(5 Pt A): 1883–1895. PubMed Abstract \| Publisher Full Text\n\nStokman G, Kors L, Bakker PJ, et al.: NLRX1 dampens oxidative stress and apoptosis in tissue injury via control of mitochondrial activity. J Exp Med. 2017; 214(8): 2405–2420. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMoore CB, Bergstralh DT, Duncan JA, et al.: NLRX1 is a regulator of mitochondrial antiviral immunity. Nature. 2008; 451(7178): 573–577. PubMed Abstract \| Publisher Full Text\n\nXia X, Cui J, Wang HY, et al.: NLRX1 negatively regulates TLR-induced NF-κB signaling by targeting TRAF6 and IKK. Immunity. 2011; 34(6): 843–53. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAllen IC, Moore CB, Schneider M, et al.: NLRX1 protein attenuates inflammatory responses to infection by interfering with the RIG-I-MAVS and TRAF6-NF-κB signaling pathways. Immunity. 2011; 34(6): 854–865. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSoares F, Tattoli I, Wortzman ME, et al.: NLRX1 does not inhibit MAVS-dependent antiviral signalling. Innate Immun. 2013; 19(4): 438–48. PubMed Abstract \| Publisher Full Text\n\nLeemans JC, Butter LM, Teske GJ, et al.: The toll interleukin-1 receptor (IL-1R) 8/single Ig domain IL-1R-related molecule modulates the renal response to bacterial infection. Infect Immun. 2012; 80(11): 3812–3820. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRoelofs JJ, Teske GJ, Bonta PI, et al.: Plasminogen activator inhibitor-1 regulates neutrophil influx during acute pyelonephritis. Kidney Int. 2009; 75(1): 52–59. PubMed Abstract \| Publisher Full Text\n\nLattenist L, Teske G, Claessen N, et al.: The lectin like domain of thrombomodulin is involved in the defence against pyelonephritis. Thromb Res. 2015; 136(6): 1325–1331. PubMed Abstract \| Publisher Full Text\n\nRamakers C, Ruijter JM, Deprez RH, et al.: Assumption-free analysis of quantitative real-time polymerase chain reaction (PCR) data. Neurosci Lett. 2003; 339(1): 62–66. PubMed Abstract \| Publisher Full Text\n\nRuijter JM, Ramakers C, Hoogaars WM, et al.: Amplification efficiency: linking baseline and bias in the analysis of quantitative PCR data. Nucleic Acids Res. 2009; 37(6): e45. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHaraoka M, Hang L, Frendéus B, et al.: Neutrophil recruitment and resistance to urinary tract infection. J Infect Dis. 1999; 180(4): 1220–1229. PubMed Abstract \| Publisher Full Text\n\nXiao TS, Ting JP: NLRX1 has a tail to tell. Immunity. 2012; 36(3): 311–312. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAdams DH, Lloyd AR: Chemokines: leucocyte recruitment and activation cytokines. Lancet. 1997; 349(9050): 490–495. PubMed Abstract \| Publisher Full Text\n\nOlszyna DP, Florquin S, Sewnath M, et al.: CXC chemokine receptor 2 contributes to host defense in murine urinary tract infection. J Infect Dis. 184(3): 301–307. PubMed Abstract \| Publisher Full Text\n\nTattoli I, Carneiro LA, Jéhanno M, et al.: NLRX1 is a mitochondrial NOD-like receptor that amplifies NF-kappaB and JNK pathways by inducing reactive oxygen species production. EMBO Rep. 2008; 9(3): 293–300. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChassin C, Goujon JM, Darche S, et al.: Renal collecting duct epithelial cells react to pyelonephritis-associated Escherichia coli by activating distinct TLR4-dependent and -independent inflammatory pathways. J Immunol. 2006; 177(7): 4773–4784. PubMed Abstract \| Publisher Full Text\n\nPhilipson CW, Bassaganya-Riera J, Viladomiu M, et al.: Modeling the Regulatory Mechanisms by Which NLRX1 Modulates Innate Immune Responses to Helicobacter pylori Infection. PLoS One. 2015; 10(9): e0137839. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRicevuti G: Host tissue damage by phagocytes. Ann N Y Acad Sci. 1997; 832(1): 426–48. PubMed Abstract \| Publisher Full Text\n\nQuinn MT, Gauss KA: Structure and regulation of the neutrophil respiratory burst oxidase: comparison with nonphagocyte oxidases. J Leukoc Biol. 2004; 76(4): 760–781. PubMed Abstract \| Publisher Full Text\n\nKors L, Butter LM, Claessen N, et al.: Dataset 1 in: NLRX1 is not involved in the host defense against Escherichia coli induced pyelonephritis. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15361.d212937\n\nKors L, Butter LM, Claessen N, et al.: Dataset 2 in: NLRX1 is not involved in the host defense against Escherichia coli induced pyelonephritis. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15361.d212938\n\nKors L, Butter LM, Claessen N, et al.: Dataset 3 in: NLRX1 is not involved in the host defense against Escherichia coli induced pyelonephritis. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15361.d212939\n\nKors L, Butter LM, Claessen N, et al.: Dataset 4 in: NLRX1 is not involved in the host defense against Escherichia coli induced pyelonephritis. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15361.d212940\n\nKors L, Butter LM, Claessen N, et al.: Dataset 5 in: NLRX1 is not involved in the host defense against Escherichia coli induced pyelonephritis. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15361.d212941" }	[ { "id": "36871", "date": "30 Aug 2018", "name": "Sylvia Knapp", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis report documents predominantly negative data by showing that NLRX1 does not seem to substantially alter the innate immune response during murine urinary tract infection. NLRX1 is a member of the Nod-like receptor family, which plays a role in mitochondrial activity and has mainly been implicated in negative regulation of anti-viral and TLR-triggered inflammation.\n\nThe report is well written, all data are explained in sufficient detail and presented well. By comparing wild type and Nlrx1-deficient mice that were infected with an uropathogenic strain of E.coli, the authors detected transient differences in the inflammatory response (chemokine) and bacterial clearance (bladder, early timepoint), without any consequence on later bacterial elimination (i.e. 48h) and progression towards pyelonephritis. Likewise, using whole blood assays (cytokine secretion and phagocytosis), NLRX1 did not confer any alterations in the inflammatory or phagocytic response to LPS or whole bacteria.\n\nThe \"conclusion sentence\" of the article sounds somewhat stronger than the conclusions drawn and discussed in the entire manuscript (including the title): stating that “NLRX1 is important in attenuating the early bacterial burden in the bladder by enhancing the local pro-inflammatory cytokine response” exaggerates the fact that only MIP2 levels were reduced in Nlrx1deficient bladders (without any consequence on MPO levels). I suggest rewording this sentence.\n\nComment to source data: Dataset 3 and 4 are mixed up (therefore I checked “partly”).\n\nIt seems that these experiments were only performed once, i.e. no replicate experiments. I suggest stating this in the methods, possibly also providing an explanation for this.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4534", "date": "11 Apr 2019", "name": "Lotte Kors", "role": "Author Response", "response": "Dear Dr. Knapp, Thank you for the thoughtful comments. We have addressed the comments and questions raised and have detailed our response below. The \"conclusion sentence\" of the article sounds somewhat stronger than the conclusions drawn and discussed in the entire manuscript (including the title): stating that “NLRX1 is important in attenuating the early bacterial burden in the bladder by enhancing the local pro-inflammatory cytokine response” exaggerates the fact that only MIP2 levels were reduced in Nlrx1deficient bladders (without any consequence on MPO levels). I suggest rewording this sentence. Reply: We agree with the reviewer that the conclusion sentence is somewhat strong compared to the overall message of the article and we therefore reworded this sentence. The conclusion in the first version was based in addition to MIP2 levels on the observation (as discussed in the third paragraph of the discussion) that at 24h the bacterial burden is increased in the NLRX1 KO bladders while the pro-inflammatory cytokine levels remain at WT level. Comment to source data: Dataset 3 and 4 are mixed up (therefore I checked “partly”). Reply: We changed the titles of the datasets. It seems that these experiments were only performed once, i.e. no replicate experiments. I suggest stating this in the methods, possibly also providing an explanation for this. Reply: The experiments were performed in vivo and ex vivo, the data shown is based on biological replicates and one animal is indicated in the figures by n. For ethical reasons these experiments were performed once." } ] }, { "id": "42570", "date": "15 Jan 2019", "name": "Sheryl Justice", "expertise": [ "Reviewer Expertise urinary tract infection", "Escherichia coli", "bacterial pathogenesis" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript by Kors et al. explores the role of the nod-like receptor X1 (NLRX1) during urinary tract infection (UTI). Given the intracellular populations during UTI, the potential role of intracellular immune regulators could be important and the investigations are understudied, making the work significant. Better understanding of the immune modulators could lead to new approaches to treat UTI. The investigators use a knock out mouse to evaluate the bacterial burden in the urinary tissues as well as the immune responses. In addition, the investigators evaluate the efficacy of phagocytosis of E. coli by granulocytes and macrophages. There are alternative conclusions that are more consistent with the published literature and better explain the data than the those provided by the investigators. The analysis of the circulating immune cells was important to exclude potential confounding differences in overall immune capacity.\n\nMajor concerns:\nThe use of 10 e8 as the inoculum is known to induce early exfoliation of the superficial bladder epithelial cells in C57Bl/6 mice (at about 16 hours), and most groups use a lower inoculum to evaluate the acute phase of UTI. The bacterial burden observed in the wild type at 24 hours is consistent with prior studies that the early exfoliation leads to establishment of a quiescent intracellular reservoir between 10e3 and 10e4. The similarity of this burden at 48 hours further supports this conclusion. In this scenario, the conclusion for the NLRX1 studies would be that the exfoliation is delayed. This would be a very interesting observation given that our understanding of the host responses that modulate exfoliation in response to infection is incomplete. Microscopic studies need to be performed to evaluate the exfoliation status to draw conclusions. Please provide the limit of detection on the graphs for the bacterial burden.\n\nThe immune response is known to induce morphological changes in UPEC that will reduce the CFU recovered. The loss of NLDX1 could alter these changes and an apparent increase in CFU could be observed, but the bacterial biomass could be similar. Given the bacterial burden of the parent, the conclusion proposed in point 1 is favored. Microscopic analysis would distinguish between these conclusions.\n\nIn the discussion, the statement that begins “In fact, NLRX1 KO mice are able to clear E. coli faster….” Is not consistent with what is known about the infection. Earlier time points (6 or 16 hours) would certainly reveal that there is a higher burden of UPEC in the WT and as such, the opposite conclusion would be made. It is highly unlikely that the burden in the WT mice would start at 10 e3 and remain there throughout the course of infection. Even if this were the case, the above conclusion would still not be supported since the burden is lower in WT mice at 24 hours. Inclusion of an earlier time point is needed.\n\nAlthough statistically significant, IL-6 changes are modest and are observed at the later stages of the acute infection. There are multiple studies that have quantified IL-6 and correlated the effect of changes in IL-6 on the infection outcome. The results should be presented in context with the published literature.\n\nInvestigation for the potential role in myeloid cell function is intriguing. However, as the investigators indicated, there are potentially other functions that may be affected. The ability of the phagocytes to kill UPEC would be an interesting test and would provide mechanistic insight into the role of NLDX1 in this system.\n\nPlease provide power calculations for the immune studies, why were fewer mice included in this portion of the study? Was the burden determined for these mice?\n\nThe dataset for the FACs analyses were not included (the cytokine data was uploaded twice). Please provide an example of the gating strategy\n\nMinor concerns:\nIn the methods for the animal experiments “human” should be “humane”.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [ { "c_id": "4535", "date": "11 Apr 2019", "name": "Lotte Kors", "role": "Author Response", "response": "Dear Dr. Justice, Thank you for the thoughtful comments. We have addressed the comments and questions raised and have detailed our response below. The use of 10 e8 as the inoculum is known to induce early exfoliation of the superficial bladder epithelial cells in C57Bl/6 mice (at about 16 hours), and most groups use a lower inoculum to evaluate the acute phase of UTI. The bacterial burden observed in the wild type at 24 hours is consistent with prior studies that the early exfoliation leads to establishment of a quiescent intracellular reservoir between 10e3 and 10e4. The similarity of this burden at 48 hours further supports this conclusion. In this scenario, the conclusion for the NLRX1 studies would be that the exfoliation is delayed. This would be a very interesting observation given that our understanding of the host responses that modulate exfoliation in response to infection is incomplete. Microscopic studies need to be performed to evaluate the exfoliation status to draw conclusions. Please provide the limit of detection on the graphs for the bacterial burden. Reply: We appreciate the reviewer’s alternative interpretation of the data and payed attention to this possibility in the second paragraph of the discussion as follows: “It is in addition possible that processes like delayed E. coli-attachment, invasion and modulated exfoliation or factor secretion of superficial bladder cells contribute to the increased presence of E. coli bacteria in NLRX1 KO bladders at 24h. Whether or not direct or indirect NLRX1-mediated modulation of bladder cells contribute to the bacterial burden has not been proven yet and warrants further study.”. We agree that microscopic studies of the superficial bladder epithelial cells to evaluate exfoliation would be interesting to add however unfortunately we used all bladder material for the study of CFU counts. Instead we added this possibility as stated above. We finally would like to stress that 24h after exfoliation effective regeneration is induced making it probably difficult to convincingly show exfoliation at 48h (1). We provided the detection limit to the figure legends. 1. Niall F. Davis and Hugh D. Flood. The Pathogenesis of Urinary Tract Infections. In: Clinical Management of Complicated Urinary Tract Infection. 2011 The immune response is known to induce morphological changes in UPEC that will reduce the CFU recovered. The loss of NLDX1 could alter these changes and an apparent increase in CFU could be observed, but the bacterial biomass could be similar. Given the bacterial burden of the parent, the conclusion proposed in point 1 is favored. Microscopic analysis would distinguish between these conclusions. Reply: The assumption that the changes by NLRX1 loss will change the immune response and as such the UPEC is able to indirectly cause an increase in CFU without affecting bacterial biomass is interesting. We however showed in our paper by the ex vivo experiments in granulocytes and monocytes and by the various cytokine and chemokine levels in the bladder that when the bacterial outgrowth is increased, the immune response is not changed in absence of NLRX1. In addition, CFU counts and not bacterial biomass is as far we know the standard to measure bacterial burden in the used UTI model. This together with our main conclusion that NLRX1 is not involved in the host defense against E. coli-induced pyelonephritis we believe that further research on morphological changes to the UPEC is beyond the scope of the article. In the discussion, the statement that begins “In fact, NLRX1 KO mice are able to clear E. coli faster….” Is not consistent with what is known about the infection. Earlier time points (6 or 16 hours) would certainly reveal that there is a higher burden of UPEC in the WT and as such, the opposite conclusion would be made. It is highly unlikely that the burden in the WT mice would start at 10 e3 and remain there throughout the course of infection. Even if this were the case, the above conclusion would still not be supported since the burden is lower in WT mice at 24 hours. Inclusion of an earlier time point is needed. Reply: This is an excellent point raised by the reviewer. We agree with the reviewer that based on other studies the bladder bacterial burden in the WT is likely increased at an earlier time point than 24h and as such the results will be interpreted differently than described in the original discussion. Therefore we revised this part of the discussion as follows: ”We found that the lack of NLRX1 is associated with an increased bacterial bladder burden at 24h. Previous studies by us (unpublished data) and others (2) show that the peak in uropathogenic E. coli outgrowth from C57BL6 mice bladders is observed before 24h. This, together with our observations, indicates that the bacterial clearance in the NLRX1 deficient bladder is delayed compared to WT”. Inclusion of earlier time points will give insights in the kinetics of UPEC burden in the WT bladder, which has as indicated already been done previously. However, adding an earlier time point retrospectively in the current study to analyze CFU kinetics is not valid since there are always intra-experimental variations in the alive inoculum size that can only be verified in retrospect (a day after the inoculum is given to the animals). This would imply that we have to repeat all the in vivo experiments which is for ethical reasons not acceptable by our institute. 2. Mulvey MA, Schilling JD, Hultgren SJ. Establishment of a persistent Escherichia coli reservoir during the acute phase of a bladder infection. Infect. Immun. 2001;69(7):4572–4579. Although statistically significant, IL-6 changes are modest and are observed at the later stages of the acute infection. There are multiple studies that have quantified IL-6 and correlated the effect of changes in IL-6 on the infection outcome. The results should be presented in context with the published literature. Reply: As can be seen from figure 2C and D and figure 4C there is no statistical significant difference in IL-6 between WT and NLRX1 deficient kidney, bladder and whole blood. We however agree that putting our results in the context with the published literature will benefit the discussion and did this in the third discussion paragraph. Investigation for the potential role in myeloid cell function is intriguing. However, as the investigators indicated, there are potentially other functions that may be affected. The ability of the phagocytes to kill UPEC would be an interesting test and would provide mechanistic insight into the role of NLDX1 in this system. Reply: We agree with the reviewer that investigating the oxidative burst of myeloid cells in response to UPEC would be interesting. However as our main message is that NLRX1 is not involved in the host defense against E. coli induced pyelonephritis we believe that adding such an experiment is beyond the scope of this paper. As an alternative we mention in the discussion that from our data it is not clear if NLRX1 plays a role in the oxidative burst in myeloid cells during UTI. Please provide power calculations for the immune studies, why were fewer mice included in this portion of the study? Was the burden determined for these mice? Reply: Based on the protocol optimized in our institute we observed that ex vivo experiments have a lower variation-coefficient than in vivo experiments and for ethical reasons we therefore included less mice. In order to increase clarity on this point we now describe the calculation in the material and method section: “To be able to reach a statistical significant effect of NLRX1 deficiency the number of 6 mice per experimental group was assessed with an unpaired t-test based on a variation coefficient of 10%, a minimal relative effect of 30%, a significance of 5% and a power of 80%.” The dataset for the FACs analyses were not included (the cytokine data was uploaded twice). Please provide an example of the gating strategy Reply: Unfortunately the FACS and Cytokine data were mixed up and we changed this. We provided the FACS gating strategy in Supplemental figure 2. Minor concerns: In the methods for the animal experiments “human” should be “humane” Reply: We changed this." } ] } ]	1	https://f1000research.com/articles/7-1197
https://f1000research.com/articles/8-750/v1	29 May 19	{ "type": "Opinion Article", "title": "Human respiratory syncytial virus methyl transferase: a potential antiviral target?", "authors": [ "Raj Kalkeri", "Govinda Bhisetti", "Nagraj Mani", "Govinda Bhisetti", "Nagraj Mani" ], "abstract": "Human respiratory syncytial virus (HRSV) causes bronchiolitis and pneumonia. The role of methyltransferase (MTase) activity of HRSV polymerase in viral replication is unknown. Literature reviews of similar viral MTases and homology- modeling of RSV MTase bound to GTP and S-adenosylmethionine (SAM) have shown sequence similarity and the conserved catalytic residues (K-D-K-E) and the SAM-binding (GXGXG) domain. Combined with the recent reports of the importance of 2’O methylation of viral RNAs in the host innate immune response evasion, and its proposed role in viral replication, HRSV MTase holds promise as a potential antiviral target. Further biological validation of HRSV MTase could facilitate the discovery of novel HRSV antivirals targeting MTase enzyme activity.", "keywords": [ "RSV", "methyltransferase", "antiviral target", "replication", "broad spectrum", "polymerase", "methylation", "MTase", "innate immunity", "homology modeling" ], "content": "\n\nHuman respiratory syncytial virus (HRSV) causes lower respiratory tract infections (bronchiolitis and pneumonia) in premature babies, young children, immuno-compromised adults and bone marrow transplant patients1,2. It is a significant risk factor for asthma, wheezing2 and progression of chronic obstructive pulmonary disease (COPD)3,4. However, the molecular pathogenesis of HRSV in asthma, wheezing and COPD is not clear. HRSV-associated infections result in significant disease and mortality. In children under 5-years of age, 33.4 million cases of HRSV-associated acute lower respiratory tract infections which result in 3.4 million hospitalizations and between 53,250 and 199,000 deaths, are reported annually worldwide5.\n\nHRSV is an enveloped RNA virus with non-segmented negative sense RNA (nsNS) genome, belonging to the Paramyxoviridae family. There are two subtypes of HRSV, subtype -A and -B, circulating in the population. HRSV causes significant health problems but at present there is no effective treatment or vaccine. The U. S. Food and Drug Administration (FDA) has approved palivizumab6, for prophylactic therapy in premature infants with less than 29-weeks of gestation and children with congenital lung or heart disease. Ribavirin, a broad-spectrum antiviral agent is the only FDA approved drug for the treatment of severe HRSV disease7. However, the use of ribavirin is limited due to the potential side effects, high cost, difficulty in administration and lack of demonstrated benefit in decreasing hospitalization and mortality8. Currently, the few drug candidates in clinical development (GS-5806, ALS-8176, EDP-938) are directed towards a limited number of viral targets (HRSV F protein, polymerase and N protein respectively)9–11. Thus, novel targets in HRSV replication are needed to address the clinically unmet medical need12.\n\n\nGuanylylation and methylation functions of HRSV polymerase\n\nHRSV produces mRNAs, which are co-transcriptionally capped (guanylylated and methylated) and polyadenylated13–15. Viral mRNA methyltransferases (MTases) catalyze the transfer of methyl groups from SAM (methyl donor) to viral mRNA caps comprised of a guanosine nucleotide linked via a unique 5’-5’ linkage. This universal process of capping and methylation is important for mRNA stability and efficient translation16. Previous analysis of the L-protein sequence of non-segmented negative strand (nsNS) RNA viruses revealed the presence of 6-conserved domains (I-VI)17. These domains were implicated in viral genome replication, transcription, mRNA 5’ guanylylation, cap methylation and 3’ polyadenylation17,18. Using computational analyses of L protein of the Mononegavirales family viruses alongside site-directed mutagenesis, the MTase domain has been previously mapped to domain VI, with a putative K-D-K-E catalytic tetrad (typical 2’-O MTase fold) and a glycine-rich motif (GxGxG) (SAM binding site)18,19. A hallmark of the SAM-dependent MTase superfamily is the conserved sequences (segregated into motifs I-X) responsible for two-conserved functions- SAM binding and MTase catalytic reaction. Motifs I, III and IV are shown to be involved in SAM binding, whereas motifs IV, VI, VIII and X play a major role in the catalytic reaction17. Ribose 2’-O MTase is typically shown to have K-D-K-E tetrad residues at its core and these residues participate in the catalytic methylation reaction.\n\nStudies with vesicular stomatitis virus (a prototype for nsNS group of viruses) model have revealed some unique features of the viral mRNA cap methylation process distinct from host18. These include (1) dual specificity of MTase activity on both the N-7 guanosine and 2’-O ribose positions encoded in a single conserved region (CR VI) of L-protein; (2) sharing of the same binding site for S-adenosylmethionine (SAM), that acts as the methyl donor20–22 (3) 2’-O methylation preceding and facilitating G-N-7 methylation23, and (4) requirement of cis-elements in viral RNA for cap methylation. Mechanisms involved in mRNA capping (guanylylation and methylation) functions of HRSV L-protein are not clearly understood.\n\n\nSequence similarity of HRSV MTase domains with other viruses\n\nAlignment of the L-protein sequences from different members of the Paramyxoviridae and other nsNS RNA viruses has demonstrated conserved residues in the MTase domains19. As previously reported, conserved motifs within the HRSV MTase domain could be predicted between amino acid sequences F1821-N2025, with catalytic tetrad at K1831, D1936, K1973, E2004 and a putative SAM-binding GxGxG….D motif at positions G1853-E-G1855-A-G1857 and D1912 in HRSV L protein. MTase catalytic tetrad (K-D-K-E) and SAM binding domains (GxGxG…D) are conserved in CR VI of HRSV and VSV (Vesicular Stomatitis Virus), suggesting a similar mechanism of cap methylation24–26.\n\n\nStructure of RSV MTase\n\nAn X-ray crystal structure of RSV MTase domain is not available in the literature. Previously, a 3.8 Å resolution structure of VSV L-protein and the methyl transferase domain using electron cryomicroscopy has demonstrated that the MTase contacts both the connector and the capping domains, without direct contact with the RNA dependent RNA polymerase (RdRp)27. However, the authors also predicted the substantial conformational change in the L protein following initial polymerization. The lack of a high resolution X-ray crystal structure of the RSV MTase is a major caveat for a structure based drug design effort.\n\nIn the absence of an X-ray crystal structure, homology modeling provides an alternative approach to model the protein structure using the crystal structure of related protein(s) where significant sequence identity/similarity exists. Sequence alignment of HRSV polymerase suggested an overall low protein sequence similarity among the members of different nsNS viruses. Similarly, low sequence homology (8–13%) was observed between HRSV MTase domain and those for which X-ray structures were available namely dengue virus NS5, vaccinia virus VP39 and E. coli RrmJ that precluded the building of a homology model of HRSV MTase. Fortunately, crystal structure of human metapneumovirus (HMPV) methyltransferase were reported recently28. HMPV is a paramyxovirus of the Pneumovirinae subfamily closely related to respiratory syncytial virus (HRSV). The MTase domains of these viruses have about 35% sequence identity and close to 60% sequence similarity allowing homology-based modeling of the RSV MTase domain. The suitability of crystal structure of human metapneumovirus (HMPV) methyltransferase for homology modeling of HRSV MTase and structure-based drug discovery needs to be further explored.\n\n\nGenetic support for RSV MTase role in viral replication\n\nRecombinant VSVs (a prototype of nsNS RNA viruses) with point mutations at methyltransferase catalytic site (rVSV-K1651A, -D1762A, and -E1833Q) were reported to be defective in cap methylation and demonstrated reduced growth in cell culture and mice29. Though SAM binding site point mutations (rVSV-G1670A, G1672A, G1674A and G4A), were attenuated in vitro, low level virulence was still observed in vivo18. In contrast, mutations in the SAM binding site of L-protein in recombinant flaviviruses and metapneumoviruses attenuated viral replication in cell culture and animal models (cotton rats and turkeys), supporting the importance of MTase for viral replication and virulence30.\n\nPreviously, HRSV transcription was shown to be independent of cap methylation, where S-adenosyl-L-homocysteine (SAH), a byproduct of MTase activity, did not affect HRSV transcription despite SAM-dependent inhibition of methylation31. However, this observation was based only upon in vitro transcription using infected cell extracts without evaluation of the quality/stability of the HRSV transcripts. For VSV (a prototype nsNS RNA virus), SAH was shown to affect the quality of mRNA (aberrant polyadenylation) without an apparent effect on transcription32,33. Thus, evaluation of the quality of transcribed viral mRNAs in the presence of SAH might reveal a previously uncharacterized effect on HRSV MTase activity. The HRSV minigenome assay provides an opportunity to address this possibility, as it is performed in cells with measurement of both transcription and translation34. Based on the model suggested for HRSV transcription15, MTase activity seems to affect late elongation or polyadenylation. However, additional studies are needed to confirm this hypothesis for HRSV MTase.\n\nRecent reports suggest that amino acid substitutions in the conserved SAM binding site and MTase domain of metapneumoviruses result in defective mRNA cap methylations and attenuate viral replication in vivo25,26. Since metapneumoviruses also belong to the Pneumovirinae subfamily as HRSV, it is possible that inhibition of HRSV MTase and SAM binding functions will negatively affect viral mRNA transcription and consequently, viral replication.\n\n\nHRSV MTase as an antiviral target\n\nSeveral small molecule inhibitors of viral MTase such as Sinefungin (SIN) and S-adenosyl-L-homocysteine (SAH) derivatives have been reported. SIN, a natural SAM-analog and a potent inhibitor of MTase, shows antiviral activity against VSV, Newcastle disease and vaccinia viruses35. Similarly, derivatives of SAH, a byproduct of mRNA cap methylation, have shown selective inhibition of MTase of dengue virus36. Key residues in HRSV MTase catalytic motif and SAM binding domain seem to be conserved between different subtypes A and B of HRSV, paramyxoviruses and negative sense non-segmented viruses except Bornaviridae family members19,37. Such a conserved sequence could provide a basis for structure-based design for pan-antiviral inhibitors targeting viral MTase26. The recently established X-ray structure of the MTase domain of HMPV might allow us to build a homology model for HRSV MTase28. This new information could facilitate structure-guided drug design.\n\nThe 2’O methylation of viral RNA is reported to be important for viral evasion of host innate immunity. The interferon-induced proteins with tetratricopeptide repeats (IFIT) are a part of the innate immune response needed to defend against viruses, recognize unmethylated mRNAs as “non-self” and target them for degradation, thus underscoring the importance of mRNA capping in replication of West Nile virus (WNV), Poxviruses, Coronaviruses, and HMPV26,30. Although additional studies are needed to confirm this phenomenon for HRSV, one could envision a potential “double whammy” effect of HRSV MTase inhibition on viral replication; on one hand the inhibition of MTase activity may attenuate or block viral transcription and replication whilst the generation of unmethylated viral mRNA caps will result in its degradation by the cellular innate immune response machinery. Due to the involvement of multiple amino acid residues for enzyme activity, targeting HRSV MTase might offer a high barrier for resistance emergence. Furthermore, mutations in these residues could attenuate interactions with SAM and/or result in reduced catalytic efficiency leading to poor replicative fitness. Mechanistic distinctions between nsNS viral MTases (using VSV as a prototype) in comparison to host MTases could enable the discovery of antivirals that selectively target viral MTases but spare host MTases thus minimizing potential toxicity23.\n\nThe high degree of sequence conservation of the HRSV MTase catalytic residues and the fundamental differences between viral and host capping mechanisms combined with the potential for the restoration of innate immune response that could specifically degrade viral mRNAs makes HRSV MTase a logical target for HRSV drug discovery efforts. Due to the sequence similarity of HRSV MTase with other members of the paramyxovidae family, it is likely that an HRSV MTase inhibitor also have activity against paramyxoviruses. Such an antiviral spectrum might be added value for empirical treatment, especially due to unavailability/delay of virus-specific diagnostics and short time to treatment initiation. HRSV MTase inhibitors should be counter screened against viral panels to determine the antiviral specificity as is the norm for antiviral drug discovery efforts during lead optimization.\n\n\nData availability\n\nNo data are associated with this article.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nHervás D, Reina J, Yañez A, et al.: Epidemiology of hospitalization for acute bronchiolitis in children: differences between RSV and non-RSV bronchiolitis. Eur J Clin Microbiol Infect Dis. 2012; 31(8): 1975–1981. PubMed Abstract \| Publisher Full Text\n\nMiller EK, Gebretsadik T, Carroll KN, et al.: Viral etiologies of infant bronchiolitis, croup and upper respiratory illness during 4 consecutive years. Pediatr Infect Dis J. 2013; 32(9): 950–955. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKokturk N, Bozdayi G, Yilmaz S, et al.: Detection of adenovirus and respiratory syncytial virus in patients with chronic obstructive pulmonary disease: Exacerbation versus stable condition. Mol Med Rep. 2015; 12(2): 3039–3046. PubMed Abstract \| Publisher Full Text\n\nSeemungal T, Harper-Owen R, Bhowmik A, et al.: Respiratory viruses, symptoms, and inflammatory markers in acute exacerbations and stable chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2001; 164(9): 1618–1623. PubMed Abstract \| Publisher Full Text\n\nNair H, Nokes DJ, Gessner BD, et al.: Global burden of acute lower respiratory infections due to respiratory syncytial virus in young children: a systematic review and meta-analysis. Lancet. 2010; 375(9725): 1545–1555. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAmerican Academy of Pediatrics Committee on Infectious Diseases; American Academy of Pediatrics Bronchiolitis Guidelines Committee: Updated guidance for palivizumab prophylaxis among infants and young children at increased risk of hospitalization for respiratory syncytial virus infection. Pediatrics. 2014; 134(2): e620–638. PubMed Abstract \| Publisher Full Text\n\nSidwell RW, Huffman JH, Khare GP, et al.: Broad-spectrum antiviral activity of Virazole: 1-beta-D-ribofuranosyl-1,2,4-triazole-3-carboxamide. Science. 1972; 177(4050): 705–706. PubMed Abstract \| Publisher Full Text\n\nTurner TL, Kopp BT, Paul G, et al.: Respiratory syncytial virus: current and emerging treatment options. Clinicoecon Outcomes Res. 2014; 6: 217–225. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMackman RL, Sangi M, Sperandio D, et al.: Discovery of an oral respiratory syncytial virus (RSV) fusion inhibitor (GS-5806) and clinical proof of concept in a human RSV challenge study. J Med Chem. 2015; 58(4): 1630–1643. PubMed Abstract \| Publisher Full Text\n\nEDP-938 clinical trial. http\n\nWang G, Deval J, Hong J, et al.: Discovery of 4'-chloromethyl-2'-deoxy-3',5'-di-O-isobutyryl-2'-fluorocytidine (ALS-8176), a first-in-class RSV polymerase inhibitor for treatment of human respiratory syncytial virus infection. J Med Chem. 2015; 58(4): 1862–1878. PubMed Abstract \| Publisher Full Text\n\nFearns R, Deval J: New antiviral approaches for respiratory syncytial virus and other mononegaviruses: Inhibiting the RNA polymerase. Antiviral Res. 2016; 134: 63–76. PubMed Abstract \| Publisher Full Text\n\nBarik S: Transcription of human respiratory syncytial virus genome RNA in vitro: requirement of cellular factor(s). J Virol. 1992; 66(11): 6813–6818. PubMed Abstract \| Free Full Text\n\nCollins PL, Karron R: Respiratory syncytial virus. In: Fields virology, 6th ed, D. M. Knipe PMH 2013. Reference Source\n\nLiuzzi M, Mason SW, Cartier M, et al.: Inhibitors of respiratory syncytial virus replication target cotranscriptional mRNA guanylylation by viral RNA-dependent RNA polymerase. J Virol. 2005; 79(20): 13105–13115. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBanerjee AK: 5'-terminal cap structure in eucaryotic messenger ribonucleic acids. Microbiol Rev. 1980; 44(2): 175–205. PubMed Abstract \| Free Full Text\n\nPoch O, Blumberg BM, Bougueleret L, et al.: Sequence comparison of five polymerases (L proteins) of unsegmented negative-strand RNA viruses: theoretical assignment of functional domains. J Gen Virol. 1990; 71(Pt 5): 1153–1162. PubMed Abstract \| Publisher Full Text\n\nLi J, Wang JT, Whelan SP: A unique strategy for mRNA cap methylation used by vesicular stomatitis virus. Proc Natl Acad Sci U S A. 2006; 103(22): 8493–8498. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFerron F, Longhi S, Henrissat B, et al.: Viral RNA-polymerases -- a predicted 2'-O-ribose methyltransferase domain shared by all Mononegavirales. Trends Biochem Sci. 2002; 27(5): 222–224. PubMed Abstract \| Publisher Full Text\n\nGrdzelishvili VZ, Smallwood S, Tower D, et al.: A single amino acid change in the L-polymerase protein of vesicular stomatitis virus completely abolishes viral mRNA cap methylation. J Virol. 2005; 79(12): 7327–7337. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLi J, Fontaine-Rodriguez EC, Whelan SP: Amino acid residues within conserved domain VI of the vesicular stomatitis virus large polymerase protein essential for mRNA cap methyltransferase activity. J Virol. 2005; 79(21): 13373–13384. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRahmeh AA, Li J, Kranzusch PJ, et al.: Ribose 2'-O methylation of the vesicular stomatitis virus mRNA cap precedes and facilitates subsequent guanine-N-7 methylation by the large polymerase protein. J Virol. 2009; 83(21): 11043–11050. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLi J, Zhang Y: Messenger RNA Cap Methylation in Vesicular Stomatitis Virus, a Prototype of Non-Segmented Negative-Sense RNA Virus. Methylation Anica Dricu,. In: Dricu PA (Ed.^(Eds).intechopen. 2012. Publisher Full Text\n\nOgino T, Banerjee AK: An unconventional pathway of mRNA cap formation by vesiculoviruses. Virus Res. 2011; 162(1–2): 100–109. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSun J, Wei Y, Rauf A, et al.: Methyltransferase-defective avian metapneumovirus vaccines provide complete protection against challenge with the homologous Colorado strain and the heterologous Minnesota strain. J Virol. 2014; 88(21): 12348–12363. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhang Y, Wei Y, Zhang X, et al.: Rational design of human metapneumovirus live attenuated vaccine candidates by inhibiting viral mRNA cap methyltransferase. J Virol. 2014; 88(19): 11411–11429. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiang B, Li Z, Jenni S, et al.: Structure of the L Protein of Vesicular Stomatitis Virus from Electron Cryomicroscopy. Cell. 2015; 162(2): 314–327. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPaesen GC, Collet A, Sallamand C, et al.: X-ray structure and activities of an essential Mononegavirales L-protein domain. Nat Commun. 2015; 6: 8749. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMa Y, Wei Y, Zhang X, et al.: mRNA cap methylation influences pathogenesis of vesicular stomatitis virus in vivo. J Virol. 2014; 88(5): 2913–2926. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZust R, Dong H, Li XF, et al.: Rational design of a live attenuated dengue vaccine: 2'-o-methyltransferase mutants are highly attenuated and immunogenic in mice and macaques. PLoS Pathog. 2013; 9(8): e1003521. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBarik S: The structure of the 5' terminal cap of the respiratory syncytial virus mRNA. J Gen Virol. 1993; 74(Pt 3): 485–490. PubMed Abstract \| Publisher Full Text\n\nRose JK, Lodish HF, Brock ML: Giant heterogeneous polyadenylic acid on vesicular stomatitis virus mRNA synthesized in vitro in the presence of S-adenosylhomocysteine. J Virol. 1977; 21(2): 683–693. PubMed Abstract \| Free Full Text\n\nHunt DM: Effect of analogues of S-adenosylmethionine on in vitro polyadenylation by vesicular stomatitis virus. J Gen Virol. 1989; 70(Pt 3): 535–542. PubMed Abstract \| Publisher Full Text\n\nTeng MN, Tran KC: Use of Minigenome Systems to Study RSV Transcription. Methods Mol Biol. 2016; 1442: 155–164. PubMed Abstract \| Publisher Full Text\n\nLi J, Chorba JS, Whelan SP: Vesicular stomatitis viruses resistant to the methylase inhibitor sinefungin upregulate RNA synthesis and reveal mutations that affect mRNA cap methylation. J Virol. 2007; 81(8): 4104–4115. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLim SV, Rahman MB, Tejo BA: Structure-based and ligand-based virtual screening of novel methyltransferase inhibitors of the dengue virus. BMC Bioinformatics. 2011; 12 Suppl 13: S24. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBujnicki JM, Rychlewski L: In silico identification, structure prediction and phylogenetic analysis of the 2'-O-ribose (cap 1) methyltransferase domain in the large structural protein of ssRNA negative-strand viruses. Protein Eng. 2002; 15(2): 101–108. PubMed Abstract \| Publisher Full Text" }	[ { "id": "49220", "date": "19 Jun 2019", "name": "Rachel Fearns", "expertise": [ "Reviewer Expertise RSV transcription and genome replication mechanisms" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis review article focuses on the methyltransferase activity of the respiratory syncytial virus (RSV) polymerase. The review covers the role of the methyltransferase in mRNA transcription and draws parallels with related viruses to describe the features of the methyltransferase. It also makes the case that inhibition of the methyltransferase activity could have a dual impact on the virus by inhibiting viral gene expression and by eliciting innate immune responses. There are a few errors that should be addressed to make the article scientifically sound. These are listed below:\nThroughout the review, the authors refer to RSV and HMPV as being members of the family Paramyxoviridae. This is no longer the case, they are now in a new family called Pneumoviridae. In addition, RSV is now formally termed human orthopneumovirus. The text should be revised to accommodate the new family name.\n\nThe authors also refer to the RSV mRNAs being modified by guanylylation, however studies by Ogino and coworkers with VSV and other rhabdoviruses have shown that the mRNA cap is added by a GDP polyribonucleotidyltransferase activity, rather than by guanylylation. Given the similarities within their capping domains, it is highly likely that RSV uses the same mechanism as the rhabdoviruses (although this has not yet been shown). Therefore, the text should be modified to remove reference to guanylyltransferase activity and a sentence or two added to explain the distinctive capping mechanism that is likely used. Likewise, the phrase \"This universal process of capping and methylation...\" should be rewritten as the processes are distinct, even though the end products are chemically equivalent.\n\nIn the description of the polymerase, the authors refer to 6 conserved domains. It is more appropriate to use the term \"regions\" rather than \"domains\".\n\nIn the section titled \"Genetic support for RSV MTase role in viral replication\" there is reference to work performed on the metapneumovirus polymerase, but this work is not cited (it is cited elsewhere in the article). The work from the Li lab should be cited here.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes", "responses": [] }, { "id": "49405", "date": "02 Jul 2019", "name": "Dirk Jochmans", "expertise": [ "Reviewer Expertise Antiviral research." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nKalkeri et al. investigate the potential of the human RSV MTase as target for antiviral drug discovery. This is a very relevant topic, as RSV remains an important human pathogen for which it has been very challenging to find potent therapies and/or vaccines. MTase of several other viruses have been investigated and the authors link this information to RSV with the hope to find new starting points for RSV drug discovery.\nIn general, we do not have the impression this review brings a lot of value. Several of the hallmark papers in the field (example Nature Communications volume 6, Article number: 8749 (2015)1 or Nature Reviews Microbiology volume 10, pages 51–65 (2012)2) describe many of the arguments of the authors in a more understandable/graphical way.\nThe value of the manuscript is also limited as it is challenging to read. It lacks a clear explanation of the different biochemical reactions involved in vRNA capping. A reaction scheme would be very helpful for the reader (like for example in Nature Reviews Microbiology volume 10, pages 51–65 (2012)2). The paper also lacks an explanation on what is SAM and how SAH is a by-product from MTase activity that has SAM-dependent inhibition.\nSimilarly, the work describes several conserved sequences and residues but an overview in a figure of these gene/protein elements is lacking while it would be extremely helpful for the reader to keep track. Without this basic information represented in a schematic way it is impossible to link the understanding of MTase from other nsNS viruses with the potential implications for RSV.\nThe text also shows several language errors including many missing occasions of ‘the’ and repeating messages like the fact that the K-D-K-E motif is a typical 2’-O MTase fold is mentioned twice under “Guanylation and methylation functions\". Also the fact that sequence similarity for homology-based modelling is important is suggested at several places in the text.\nAlso, some relevant arguments remain unexplained and thus difficult to understand. For example “MTase activity seems to affect late elongation or polyadenylation” or “Due to the involvement of multiple amino acid residues for enzyme activity, targeting HRSV MTase might offer a high barrier for resistance emergence.”.\nTwo aspects of the paper are certainly interesting. These are the parts on the genetic support for RSV MTase role in viral replication and the importance of 2’O methylation on viral RNA. They clearly represent the challenges and opportunities for finding novel MTase inhibitors.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-750
https://f1000research.com/articles/8-1677/v1	23 Sep 19	{ "type": "Opinion Article", "title": "BioHackathon series in 2013 and 2014: improvements of semantic interoperability in life science data and services", "authors": [ "Toshiaki Katayama", "Shuichi Kawashima", "Gos Micklem", "Shin Kawano", "Jin-Dong Kim", "Simon Kocbek", "Shinobu Okamoto", "Yue Wang", "Hongyan Wu", "Atsuko Yamaguchi", "Yasunori Yamamoto", "Erick Antezana", "Kiyoko F. Aoki-Kinoshita", "Kazuharu Arakawa", "Masaki Banno", "Joachim Baran", "Jerven T. Bolleman", "Raoul J. P. Bonnal", "Hidemasa Bono", "Jesualdo T. Fernández-Breis", "Robert Buels", "Matthew P. Campbell", "Hirokazu Chiba", "Peter J. A. Cock", "Kevin B. Cohen", "Michel Dumontier", "Takatomo Fujisawa", "Toyofumi Fujiwara", "Leyla Garcia", "Pascale Gaudet", "Emi Hattori", "Robert Hoehndorf", "Kotone Itaya", "Maori Ito", "Daniel Jamieson", "Simon Jupp", "Nick Juty", "Alex Kalderimis", "Fumihiro Kato", "Hideya Kawaji", "Takeshi Kawashima", "Akira R. Kinjo", "Yusuke Komiyama", "Masaaki Kotera", "Tatsuya Kushida", "James Malone", "Masaaki Matsubara", "Satoshi Mizuno", "Sayaka Mizutani", "Hiroshi Mori", "Yuki Moriya", "Katsuhiko Murakami", "Takeru Nakazato", "Hiroyo Nishide", "Yosuke Nishimura", "Soichi Ogishima", "Tazro Ohta", "Shujiro Okuda", "Hiromasa Ono", "Yasset Perez-Riverol", "Daisuke Shinmachi", "Andrea Splendiani", "Francesco Strozzi", "Shinya Suzuki", "Junichi Takehara", "Mark Thompson", "Toshiaki Tokimatsu", "Ikuo Uchiyama", "Karin Verspoor", "Mark D. Wilkinson", "Sarala Wimalaratne", "Issaku Yamada", "Nozomi Yamamoto", "Masayuki Yarimizu", "Shoko Kawamoto", "Toshihisa Takagi", "Shuichi Kawashima", "Gos Micklem", "Shin Kawano", "Jin-Dong Kim", "Simon Kocbek", "Shinobu Okamoto", "Yue Wang", "Hongyan Wu", "Atsuko Yamaguchi", "Yasunori Yamamoto", "Erick Antezana", "Kiyoko F. Aoki-Kinoshita", "Kazuharu Arakawa", "Masaki Banno", "Joachim Baran", "Jerven T. Bolleman", "Raoul J. P. Bonnal", "Hidemasa Bono", "Jesualdo T. Fernández-Breis", "Robert Buels", "Matthew P. Campbell", "Hirokazu Chiba", "Peter J. A. Cock", "Kevin B. Cohen", "Michel Dumontier", "Takatomo Fujisawa", "Toyofumi Fujiwara", "Leyla Garcia", "Pascale Gaudet", "Emi Hattori", "Robert Hoehndorf", "Kotone Itaya", "Maori Ito", "Daniel Jamieson", "Simon Jupp", "Nick Juty", "Alex Kalderimis", "Fumihiro Kato", "Hideya Kawaji", "Takeshi Kawashima", "Akira R. Kinjo", "Yusuke Komiyama", "Masaaki Kotera", "Tatsuya Kushida", "James Malone", "Masaaki Matsubara", "Satoshi Mizuno", "Sayaka Mizutani", "Hiroshi Mori", "Yuki Moriya", "Katsuhiko Murakami", "Takeru Nakazato", "Hiroyo Nishide", "Yosuke Nishimura", "Soichi Ogishima", "Tazro Ohta", "Shujiro Okuda", "Hiromasa Ono", "Yasset Perez-Riverol", "Daisuke Shinmachi", "Andrea Splendiani", "Francesco Strozzi", "Shinya Suzuki", "Junichi Takehara", "Mark Thompson", "Toshiaki Tokimatsu", "Ikuo Uchiyama", "Karin Verspoor", "Mark D. Wilkinson", "Sarala Wimalaratne", "Issaku Yamada", "Nozomi Yamamoto", "Masayuki Yarimizu", "Shoko Kawamoto", "Toshihisa Takagi" ], "abstract": "Publishing databases in the Resource Description Framework (RDF) model is becoming widely accepted to maximize the syntactic and semantic interoperability of open data in life sciences. Here we report advancements made in the 6th and 7th annual BioHackathons which were held in Tokyo and Miyagi respectively. This review consists of two major sections covering: 1) improvement and utilization of RDF data in various domains of the life sciences and 2) meta-data about these RDF data, the resources that store them, and the service quality of SPARQL Protocol and RDF Query Language (SPARQL) endpoints. The first section describes how we developed RDF data, ontologies and tools in genomics, proteomics, metabolomics, glycomics and by literature text mining. The second section describes how we defined descriptions of datasets, the provenance of data, and quality assessment of services and service discovery. By enhancing the harmonization of these two layers of machine-readable data and knowledge, we improve the way community wide resources are developed and published. Moreover, we outline best practices for the future, and prepare ourselves for an exciting and unanticipatable variety of real world applications in coming years.", "keywords": [ "BioHackathon", "Bioinformatics", "Semantic Web", "Web services", "Ontology", "Databases", "Semantic interoperability", "Data models", "Data sharing", "Data integration" ], "content": "Introduction\n\nBig data in the life sciences - especially from ‘omics’ technologies - is challenging researchers with scalability concerns in terms of computational and storage needs, while at the same time, there is also a stronger drive towards the promotion of open data including the sharing of analyses and their outputs. Consistent with this, the \"Open Data Charter\" issued by the 2013 G8 summit meeting states that the release of high-value open data is important for improving democracies and encouraging innovative reuse of data. Experimental results including genome data, as well as research and educational activities, are recognized as of high value in the Science and Research category of the Charter. To fully utilize open data in life sciences, semantic interoperability and standardization of data are required to allow innovative development of applications.\n\nDuring the 6th and 7th NBDC/DBCLS BioHackathons in 2013 and 2014, which were hosted by the National Bioscience Database Center (NBDC) and the Database Center for Life Science (DBCLS) in Japan, we focused on the improvement of Resource Description Framework (RDF) data for practical use in biomedical applications by developing guidelines, ontologies and tools especially for the genome, proteome, interactome and chemical domains. Also, to host these data effectively, we explored best practices for representing dataset metadata, as well as assessing the capabilities of triple stores and the quality of service of endpoints. The BioHackathon 2013 was held in Tokyo and BioHackathon 2014 was held in Miyagi. Both were sponsored by the NBDC and the DBCLS in the series of NBDC/DBCLS BioHackathons1–4, which bring together database providers and bioinformatics software developers to make their resources integrable in effective ways.\n\n\nImprovement and utilization of RDF data in life sciences\n\nPublishing data based on the RDF model and its serialization formats (e.g. Turtle), along with relevant biomedical ontologies, is becoming widely accepted within the bioinformatics community5–9 as a way of serving semantically annotated data. In this section, we describe recent developments in RDF standardization for the genomics, proteomics, glycomics, chemoinformatics and text-mining domains.\n\n\nGenomic information\n\nGenome data is a key component in modern life sciences as it serves as a hub for data integration. In the previous BioHackathons, we have developed ontologies, such as the Feature Annotation Location Description Ontology (FALDO)10 and the Genomic Feature and Variation Ontology (GFVO)11, and produced RDF data from heterogeneous datasets for integrated databases and applications. In this section, we describe how we modeled genomic annotations and related resources in RDF and ontologies.\n\nDuring the BioHackathon 20124, it was recognized that a common schema ontology was desirable for the Semantic Web integration of sequence annotation across multiple databases. In depth group discussions including bioinformatics software developers and major database representatives identified common core needs in defining locations on biological sequences (both nucleic acids and proteins). This produced a draft specification for the Feature Annotation Location Description Ontology (FALDO), and proof of principle data conversion tools. This work continued at the BioHackathon 2013, with a specific focus on ensuring that all the existing annotations in the International Nucleotide Sequence Database Collaboration (INSDC)12 feature tables could be converted into RDF triples using FALDO, as well as standardizing the coordinate system, and making sure that the starts of features are biologically sensible i.e. the start value is numerically higher than the end for genes located on the reverse strand. Subsequently, in May 2014, DBCLS organized a closed meeting, the RDF Summit, where a small group of developers from DBCLS, DNA Data Bank of Japan (DDBJ), Swiss Institute of Bioinformatics (SIB), European Bioinformatics Institute (EBI) and Stanford gathered to standardize the RDF representation of genomic annotations. The group agreed to use the FALDO ontology (see the section below) for annotating the coordinates of genomic annotations and to represent gene/transcript/exons in RDF. As a result, the RDF model of DDBJ, Ensembl13 and TogoGenome9 are now aligned such that common SPARQL queries can retrieve sequence annotations from these distinct data sources interoperably.\n\nAfter defining a common RDF model to represent the INSDC feature tables, one of the major remaining needs was to standardize the RDF representation of genome variations, which was discussed during the BioHackathon 2014.\n\nA group from EBI, DBCLS and Tohoku University surveyed existing databases that represent clinical annotation of variants. National Center for Biotechnology Information (NCBI) ClinVar14 provides information on the relationships between human genetic variation and phenotypes along with supporting evidence; Online Mendelian Inheritance in Man (OMIM)15 provides relationships between genes and disease; Leiden Open Variant Database (LOVD)16 provides gene variants related to colon cancer; Human Gene Mutation Database (HGMD)17 is commercial but widely used; Thomson Reuters Gene Variant Database (GVDB) is also a commercial database. Tohoku Medical Megabank had a license to jointly develop the RDF version of the GVDB with Thomson Reuters and they completed the initial version to test queries like \"find shared variations among diseases\" and \"find related variations from a specific disease\". In parallel, the EBI group started to convert Ensembl variation data into RDF in which an \"allele\" is related to \"gene_variant\", \"sequence_alteration\" and \"regulatory_region_variant\" instances in the sequence ontology (SO), and its location is represented by means of a FALDO region [Figure 1].\n\nThe H-invitational database (H-InvDB)18,19 group developed RDF data and an ontology for their database covering ncRNA annotations. During the Biohackathon 2013, the RDF version of the H-InvDB was expanded and its ontology was published including recent advancement in understanding of non-coding RNA (ncRNA) function. To improve descriptions of the functional relationships between coding transcripts and ncRNA, links between transcripts in H-InvDB and two major RNA databases, Rfam20 and miRBase21, were added. For miRBase, interactions between miRNA and transcripts were predicted using TargetScan22. For both of these databases, new classes were defined in the ontology to describe interaction events, such as binding between a transcript and a miRNA. At the BioHackathon 2014, the group tried to incorporate variant information into the RDF data.\n\nThere was discussion of how to represent gene names and chromosome Uniform Resource Identifiers (URIs). For gene names, it is recommended to use rdfs:label and dc:identifier for primary gene IDs and use skos:altLabel for gene synonyms. However, it is not mandatory because gene IDs are not always available, depending on the source of information. As for chromosome URIs, it would be useful if the bioinformatics community could agree to a common URI for each chromosome and version (e.g. human chromosome 19 in the GRCh38 assembly). However, we could not reach an agreement at the BioHackathon as it seemed to be impractical to cover every sequence assembly of all species, individuals, cells and samples in an unified manner as drafted in the RDF summit. In this section, we describe the current situation and proposals relating to this issue.\n\nUniversal Biological Sequence ID (UBSID). An essential step in the merging of datasets is relating primary identifiers i.e. any data can be joined if they contain the same identifiers. Therefore, all databases can be joined as fully connected Linked Data if appropriate universal identifiers are consistently used. To date, molecular biology has mainly developed around the Central Dogma concept in which higher levels of annotation (transcripts, proteins) are related to the underlying genomic sequence. Genes, as well as protein binding motifs and other features such as SNPs, can be related to DNA sequences, as can the transcriptome and proteome. Therefore, much of modern molecular biology data can in principle be related if the underlying nucleotide sequences are used as the basis for identifiers. However, the use of sequences per se as identifiers has several problems: for example, a sequence can be extremely long (e.g. human chromosome I), or very short (e.g. the location of a SNP), there can be multiple sequences that are highly similar or identical as in multi-copy paralogs, and a sequence feature can be on the sense or antisense strand. In order to overcome these problems, a universal sequence-based identifier scheme should incorporate position information, reference sequence information, the actual sequence (when there are differences, such as mutations, from the reference), strand information, and in addition, it would be ideal if all of such information is expressed as a short, human-comprehensible identifier. By using reference-based compression of DNA sequences based on offset and run-length encoding, the sequence can be expressed just by the mismatching positions and this can form the basis for an identifier system. Therefore, the G-language group proposed a Universal Biological Sequence ID (UBSID) to enable this encoding. For example, the human APOE mRNA sequence is encoded as <http://rest.g-language.org/ubsid/ubsid2seq/hg19-chr19:045409882+A42:=43-1092=193-580=718:> as a URI in the G-language REST service.\n\nIdentifiers used in the DDBJ and TogoGenome RDF. After the BioHackathon 2012, a group from DBCLS and DDBJ developed an ontology which can capture semantically the data model of the INSDC, such as the records from GenBank, DDBJ and ENA, with restrictions on terms used in the feature table and qualifier key-values. A converter for INSDC and RefSeq23 entries to RDF was developed based on the ontology, and the RDFized data is used in the TogoGenome application. TogoGenome integrates information on genes, proteins, organisms, phenotypes and environments. Because the genes in TogoGenome are currently extracted from INSDC and RefSeq records, the URI for each annotation is constructed as a fragment using Identifiers.org URIs in the form: <http://identifiers.org/[insd c or refseq]/[entry_id]#[fragment]>. For example, the human APOE gene on chromosome 19 in the RefSeq record NC_000019.10 is internally represented in TogoGenome as <http://identifiers.org/refseq/NC_000019.10#feature:44905782-44909393:1:gene.1424> and the information can be accessed at <http://togogenome.org/gene/9606:APOE> where 9606 is the taxon ID corresponding to human in the NCBI Taxonomy database and APOE is the gene name used in the record. This approach is slightly different from the proposed UBSID model which encodes sequence alignment with comments but can distinguish the source of information and feature types annotated in the INSDC/RefSeq record. The location of each gene and exons in TogoGenome RDF are described by the FALDO ontology.\n\nIdentifiers used in the Ensembl RDF. Ensembl generates their own IDs for genes, transcripts and exons in their database. For example, the human APOE gene is given an ID of ENSG00000130203, which encodes five transcripts (one of them is ENST00000252486) and one of the exons of this transcript is ENSE00003577086. It is natural to use these IDs when constructing URIs for the RDF dataset. In the 2014 development version of the Ensembl RDF, the human APOE gene is indicated as <http://rdf.ebi.ac.uk/resource/ensembl/ENSG00000130203> within a graph identified as <http://rdf.ebi.ac.uk/dataset/ensembl/77/9606> for the human genome dataset in the Ensembl release 77. The location of this gene on human chromosome 19 is designated by <http://rdf.ebi.ac.uk/resource/ensembl/77/chromosome:GRCh38:19:44905754-44909393:1>. The strategy to generate unique URIs for each annotation in Ensembl is different from that employed by DDBJ/INSDC and TogoGenome, which all share both the same RDF model and use the FALDO ontology to describe the actual coordinates of annotations (e.g. genes and exons) on a chromosome. Thus at present further work is needed before all these providers are completely consistent and interchangeable.\n\nTo facilitate more accurate and deeper integration of data, it is important to standardize metadata accompanying DNA sequences, orthologous gene relationships among organisms, phenotypic properties of organisms, inter-species and organism-environment interactions including host-pathogen relationships. We describe some of these efforts now.\n\nMetadata on samples. DDBJ, EBI and NCBI are jointly hosting the BioSample database as an international collaboration. In this resource, metadata are accumulated on the samples from which DNA sequence in the INSDC database was collected and/or on which other research projects were conducted. The metadata includes species, type of samples (cell types etc.) and phenotypic or environmental information, and therefore it is valuable for data integration if the metadata is available as RDF. A group from DDBJ generated an RDF version of BioSample metadata during the BioHackathon 2014, using as a starting point 14,362 entries stored in the DDBJ BioSample database in XML format.\n\nIn addition, existing terminologies and ontologies for geological, archeological and morphological data were explored during the 2014 BioHackathon. For example, there are several resources for geolocations such as W3C Geospatial Ontologies, GeoRSS, GeoNames and Global Biodiversity Information Facility (GBIF). The National Aeronautics and Space Administration (NASA) has developed the Global Change Master Directory (GCMD) and the Semantic Web for Earth and Environmental Terminology (SWEET) which can be used to describe archaeological time scales. For morphology, the Foundational Model of Anatomy (FMA)24, Anatomy Reference Ontology (AEO)25, Vertebrate Skeletal Anatomy Ontology (VSAO)26 and other domain specific ontologies27 were surveyed. These ontologies are essential for encoding RDF data in environmental biology, such as biodiversity and biomolecular archeology. As a case study, a group developed a semantic resource with information about corals by integrating taxonomic, genomic, environmental, disease and coral bleaching information.\n\nOntologies for integration of microbial data. Within the field of microbiology, genomic and metagenomic data are expanding rapidly due to advances in next generation sequencing technologies. To effectively analyze these huge amounts of data, it is necessary to integrate various microbial data resources available on the Internet. Orthology can play an important role in summarizing such data by grouping corresponding genes across different organisms, and by annotating genes by transferring knowledge from highly curated model organism to newly sequenced genomes. Therefore RDF models were developed for representing the orthology data stored in the Microbial Genome Database for Comparative Analysis (MBGD)8, and these were used to construct an RDF version of MBGD. This also required the development of the OrthO ontology28 for representing orthology and aligning concepts with the existing OGO ontology29, with additional definitions mapped from OrthoXML30. Orthology RDF data can now be linked with other databases published as RDF such as UniProt31, allowing the integrated dataset to be queried using SPARQL. When searching these data, ontologies can be utilized to specify complex search conditions. To assist making such precise queries, the Microbial Phenotype Ontology (MPO) was developed for describing microbial phenotypes such as microbial morphology, growth conditions, biochemical or physiological properties. During the hackathon, the ontology was updated to comply with a better classification of the hierarchical (is-a) and partonomical (part-of) structure. In addition, the Pathogenic Disease Ontology (PDO) was developed to describe pathogenic microbes that cause diseases in their hosts. An RDF dataset that describes pathogenic information relating to each microbial genome sequence was created using the PDO. Since the genes within these genomes are connected to the ortholog information in the MBGD ortholog database, it is possible to calculate the set of orthologous gene groups that is enriched in the disease related microbes.\n\nKnowledge extraction of factors related to diseases. Information and knowledge of the relationships between genes/ mutations/ lifestyle/ environment and diseases is required in order to predict the risk of a disease and for prognosis after the onset of a disease. In practice, it will also be necessary to collect individual lifestyle and environmental profiles as well as personal genetic data such as genome sequences to allow such predictions for individual people. The necessary underlying relationships are often described in the literature, but are not yet systematically collected in a database. To extract these relationships from the literature, there are two key steps that need to be addressed. First, entities must be annotated automatically using text mining software and, second, these annotations must be represented in a curation interface to allow confirmation that the information has been extracted accurately. Genes, genetic variants, diseases, environmental factors and lifestyle factors are the entity types that need to be annotated on the corpus. Existing software for extracting genes (e.g. GNAT32, GenNorm33 etc.), mutations (e.g. tmVar34, MutationFinder35 etc.) and diseases (e.g. BANNER36 with disease model) are openly available, along with existing datasets such as BioContext37 and EVEX DB38. Before environmental factors and lifestyle factors can be extracted systematically it is necessary to decide on a controlled vocabulary (whether existing or not) to represent them. Pregnancy Induced Hypertension (PIH) was chosen as a case study and 86 relevant open access PubMed Central articles identified. It was possible to extract genes from 32 of these articles using the BioContext dataset, while the other 54 articles were published more recently than BioContext. Attempts were made to extract mutations from 86 articles. For lifestyle and environmental factors, controlled vocabularies were collected in preparation for entity recognition. After obtaining all the entities in the 86 articles, they were curated using interfaces such as PubAnnotation39, and the curated relationships represented as an RDF graph.\n\nGenome annotations have historically been represented and distributed in non-standard domain-specific data formats (e.g., INSDC, GFF3, GTF). The data formats themselves often include implicit semantics, making automatic interpretation and integration of the data with other resources challenging. Therefore, tools to convert those data into RDF and ontologies to support semantic representation of data need to be developed. BioInterchange is a tool to convert those file formats into RDF and was originally developed in the BioHackathon 20124, with its functionalities and ontologies being enhanced over successive hackathons. Other tools for high-throughput data processing of Sequence Alignment/Map format (SAM), Binary SAM format (BAM)40, Variant Call Format (VCF)41, Genome Variation Format (GVF)42, Header-Dictionary-Triples (HDT)43 files have also been developed and a middleware to enable SPARQL queries directly against these huge files on-the-fly for scalability was explored and results were incorporated into integrated semantic genome databases such as TogoGenome and MicrobeDB.jp.\n\nUtilization of domain specific data formats in semantic web. In BioHackathon 2013, VCF2RDF was developed and subsequently published as a Ruby program to convert VCF files into RDF, which represent positions in FALDO and alleles in its own ad hoc ontology terms. The resulting RDF data was loaded into Fuseki and queries were tested in the Jena framework, taking three minutes for the cow genome on a laptop to plot quality scores of variant calls for a million base pairs. During BioHackathon 2014, a group developed middleware to interpret SPARQL queries against SAM/BAM/VCF files on the fly. The first implementation was prototyped in JRuby so that the Java library for samtools can be used in a Ruby program. The resulting application, VCFotf, is packaged as a Docker image that serves a query interface on the Web page. Also, another implementation (sparql-vcf) was developed with Jena for improving query execution time, in which Jena property functions are used to introduce a \"special predicate\" which accelerates search performance; however, this ‘boutique’ query violates the SPARQL standard.\n\nUse of compressed RDF for large scale genomic data. BioInterchange was used in a Genomic HDT project as a feasibility study to convert a variety of genomic data files (e.g. GVF) containing coordinate-annotated genomic features into an ontology-annotated RDF representation. The RDF data file is then processed into an RDF/HDT file, which is a compressed, indexed, and queryable data archive. Using Ensembl's human somatic variation data (81MB, 9MB gzipped), it was found that the RDF/HDT archive is only 20MB (1.5M triples; 15MB data + 5MB index), which is a significant reduction from the 313MB RDF N-triples representation. A JSON RESTful API was made available using Sinatra to provide access to the RDF/HDT file, and this allowed a demonstration of genome-based browsing of the RDF/HDT data file using the JBrowse genome browser.\n\nIntegrated semantic genome databases. TogoGenome9 was developed to integrate heterogeneous biomedical data using Semantic Web technologies. This utilizes the representation of genomic data in the standard RDF format, enabling interoperation with any other Linked Open Data (LOD) around the world. To support these efforts we have collaborated with DDBJ, UniProt, and the EBI RDF group to develop ontologies for representing locations and annotations of genome sequences and used these developments for all prokaryotic genomes and, later, eukaryotic genomes. To complement the above work we developed ontologies for taxonomies, phenotypes, environments and diseases related to organisms, so enabling faceted browsing of the entire datasets. Every TogoGenome report page is made up of modular components called TogoStanza, which is a generic framework to generate Web components querying SPARQL endpoints and rendering them as HTML elements. Stanzas are re-usable modules which can be shared and embedded easily into other databases, and which have been developed in collaborations with MicrobeDB.jp, MBGD8 and CyanoBase44, resulting in over 100 TogoStanzas being available so far.\n\nVisualization of semantic annotations in JBrowse. JBrowse45 was used by several projects within the BioHackathon as a demonstration platform. JBrowse running on top of the SPARQL endpoints, e.g. TogoGenome or a prototype InterMine46 endpoint, or from indexed files produced by GenomicHDT, were comparable in performance with typical RDB back-ended settings. In addition, an unusual use of JBrowse was to view text instead of DNA sequence, with the annotation viewed being the output of natural language processing.\n\nTogoGenome: JBrowse was extended to support the TogoGenome's SPARQL endpoint as a data source to retrieve and visualize genes on a chromosomal track (Figure 2). This enhancement is already merged into the official JBrowse release since version 1.10 in 2013. SPARQL queries are customizable in the JBrowse configuration file as long as they return start, end, strand, type (label), uniqueID and parentUniqueID of the annotation objects in a given range within a sequence. When scrolling to neighboring regions, the performance is good enough for browsing.\n\nInterMine: Representatives from the InterMine project46 produced proof-of-concept demonstrations of semantic extensions to InterMine data-warehouses. These included on the one hand a draft of how to model InterMine data as linked data, producing both an ontology of relationships and triples that conform to that ontology, and on the other hand a draft of a very limited SPARQL engine capable of operating on an InterMine data source directly. Together these investigations indicate that given some development effort, it is likely that significant progress can be made to integrating InterMine into the semantic web. An area that needs work, and is receiving attention, is the production of stable URIs for InterMine entities. In addition to this, work was done to implement a simple adaptor allowing, as described above, JBrowse to request data directly from InterMine RESTful web services.\n\nText-mining: In the community of BioHackathon, text mining resources were developed around PubAnnotation, a public repository of literature annotation data sets. Usually text mining requires its own set of tools, e.g. viewers or editors. However, an interesting experiment was carried out during BioHackathon 2013 and 2014 to use JBrowse as a viewer of text annotation data. The idea behind the experiment was that both genomic data and text data are represented as character sequences, and that annotations of both type of data are attached to specific regions on the sequences. A simple script was developed to convert annotations in PubAnnotation to JBrowse format, and it was observed that text annotations can be nicely viewed in JBrowse. The result raises the possibility of further interoperability between tools for genomics and text mining.\n\n\nProteomics, metabolomics and glycomics information\n\nIn addition to genomic information, advancements in developing ontologies and RDF datasets for proteins, metabolites, and glycans were made during the hackathons. It took several years to design standard data models as a community agreement and to convert existing resources into RDF by adding semantics, and the BioHackathons have successfully facilitated the efforts of domain experts.\n\nThe European Bioinformatics Institute’s (EBI) SIFTS \"Structure Integration with Function, Taxonomy and Sequences\" resource provides regularly updated residue-level mappings between UniProt and PDB entries47. SIFTS has been distributed in Comma Separated Values (CSV) and Extensible Markup Language (XML) formats. Like many other proteome-related databases, SIFTS uses the classical protein chain ID specified by the author. However, in 2016, the Worldwide Protein Data Bank (wwPDB) will abolish the conventional PDB format and instead will distribute RDF/XML based on the PDB exchange dictionary / macromolecular Crystallographic Information Format (PDBx/mmCIF) [PDBx/mmCIF]. At the same time wwPDB will start assigning protein chain identifiers, which will also be encoded as URIs in the wwPDB/RDF.\n\nDuring the BioHackathon, an RDF version of SIFTS (RDF-SIFTS) was designed and implemented to provide residue-to-residue correspondence between PDB and UniProt entries in RDF48. RDF-SIFTS links both the protein chain ID assigned by authors and the one assigned by wwPDB to SIFTS. RDF-SIFTS uses existing ontologies of PDB, UniProt, EMBRACE Data and Methods (EDAM)49 as well as FALDO, and resources are linked to Identifiers.org50 URIs.\n\nThe University of Tokyo Proteins (UTProt)51 is a project that is collecting and building RDF to support interactome linked data. During the BioHackathon, the UTProt group extended RDF-SIFTS to cover intermolecular interactions, and this resulted in six billion triples including, for each pair of residues in the interacting surfaces, their separation distance. This resource will be useful for analysis of structure and sequence in proteomics and interactomics. Serialization Ruby code, RDF-SIFTS maker, is available through GitHub as open source software which can be used to convert new release of SIFTS data from EBI.\n\n\"Omics\" technologies are primarily aimed at the universal detection of genes (genomics), mRNAs (transcriptomics), proteins (proteomics) and metabolites (metabolomics) in a specific biological sample. Proteomics and metabolomics in particular have gained a lot of attention in recent years due the possibility of studying reactions, post-translational modifications, and pathways52. The proteomics community has been working for more than ten years in the standardization of file formats and proteomics data53. Different XML-based file formats and open-source libraries have been released to handle proteomics data from spectra to quantitation results54–56.\n\nIn contrast, metabolomics is a relatively new \"omics\" field where the standardization of exchange formats is difficult, due to the variety of measurement methodologies ranging from nuclear magnetic resonance (NMR) spectroscopy to a variety of mass spectrometers (MS). Moreover, currently no single system can provide enough resolution to measure the entire set of small molecules within a biological sample; instead, data from multiple systems are combined to gain more comprehensive coverage, for instance combining Liquid Chromatography (LC), Gas Chromatography (GC), and Capillary Electrophoresis (CE) separation prior to analysis in a mass spectrometer. Recently, the mzTab data exchange format was introduced by the Human Proteome Organization (HUPO) Proteomics Standards Initiative, as a standardized format to report both qualitative and quantitative metabolomics and proteomics experiments in a simple tabular format57. In BioHackathon 2014, a Perl library was developed to standardize the metabolomics data obtained from MasterHands software58. MasterHands is a proprietary software for the analysis of CE-MS-based metabolomics used in the Institute for Advanced Biosciences, Keio University, and at Human Metabolome Technologies Inc. The library allows the annotation of KEGG compound information using the KEGG REST API, and also allows the annotation of Reactome and MetaCyc information.\n\nIn the age of systems biology and data integration, proteomics data represent a crucial component to understand the “whole picture” of life. In this context, well-established databases for proteomics data include the Global Proteome Machine Database (GPMDB), PeptideAtlas, ProteomicsDB, and the Proteomics Identification (PRIDE) database among others59. In addition, at BioHackathon 2014, the \"omics\" group worked on the standardization to RDF of different web services and APIs for proteomics and protein expression data. The GPMDB2RDF and PRIDE2RDF library allow the export of expression data from the GPMDB Database60 and PRIDE Database61 respectively. The development of a standard interface for providing protein expression data will allow, in the future, exchange and proper reuse of public proteomics data. To this end, the \"omics\" group in the BioHackathon 2014 made the first steps towards the development of the ProteomeXchange Interface (PROXI) for protein expression data exchange59.\n\nThe glycoscience group participated in a satellite BioHackathon in Dalian, China, in parallel to the GLYCO 22 Meeting held June 23–28, 2013. Although a preliminary RDF format was developed at the previous BioHackathon in 201262, there was a need to address not only glycan structures (sequences) but also supporting experimental data, the biological source of the sample analyzed, and publication information. Therefore, during BioHackathon 2013, a formal ontology to represent these features, as well as the glycan structures to which they relate, was discussed. The aim of the GlycoRDF group was to define a standard RDF representation, in the form of an ontology by integrating features from existing ontologies where possible and creating new classes and relationships where needed.\n\nAs it would be impossible, in a week, to create an ontology that could cover the full spectrum of glycomics information and experimental data, it was decided that the group would limit the first version to the data that currently exists in glycomics databases. On the other hand, the developers also attempted to define the ontology so that it could be easily extended with additional predicates and classes if needed, in case more data or more glyco-related databases utilize the proposed RDF format. As a result, by the end of BioHackathon 2013, the first version of the GlycoRDF ontology was agreed upon and is currently available at the GlycoRDF repository at 63. In 2014, work progressed to the point where all glyco-scientists who attended previous BioHackathons had now generated GlycoRDF-formatted versions of their databases. The updated list of these databases are listed and documented on the GlycoRDF repository.\n\nEntities can be classified based on a variety of features, such as their function(s), structures/sub-structures, or chemical properties. For example, genes and proteins are independently classified based on their functions, role, and cellular location, organized by the Gene Ontology (GO)64. At the same time, gene and proteins are also classified based on their conserved partial substructures, such as protein domains in Pfam. ChEBI65 classifies chemical substances by their overall functions (ChEBI role ontology) and by their partial structures (ChEBI molecular structure ontology). For enzymes, their overall functions are classified by the Enzyme List of International Union of Biochemistry and Molecular Biology, often referred to as the Enzyme Commission (EC) numbers66. To date, however, there have been no standard ways to classify enzymes based on the partial structures of their enzymatic reactions. Therefore, during BioHackathon 2013 we discussed the development of an ontology that deals with the partial structures of enzymatic reactions, i.e. substrate-product pairs derived from reaction equations. This led to the Enzyme Reaction Ontology for annotating Partial Information of biochemical transformation (PIERO) being published in 201467. In BioHackathon 2014, we had further discussions to refine the PIERO data to establish the PIERO Ver0.3 Schema. This ontology was later used in de novo metabolic pathway reconstruction analysis68 and for ortholog predictions69.\n\n\nText mining and question-answering\n\nIn contrast to molecular resources, extraction and utilization of knowledge represented in the literature is still in progress. As an infrastructure, it is proposed to have a common open platform for sharing text annotations resulting from manual curation and various natural language processing (NLP) techniques. NLP methods are also applied to derive a SPARQL query from natural language.\n\nText mining is becoming an increasingly common component of biological curation pipelines and biological data analysis, and as such there is increasing demand for both text that has been automatically annotated with natural language processing tools, and annotated document resources that can be used in development and evaluation of those tools. This demand in turn leads to a need for standard, interoperable representations for annotations over documents. Several proposals for general linguistic annotation representations have been made70, including ones specifically for biomedical text annotation representations71,72, as well as data models underpinning standard modular architectures such as Unstructured Information Management Architecture (UIMA)73. However, these approaches have not been adapted to the Semantic Web. Recently, the Open Annotation Core Data Model has been proposed to enable interoperable annotations on the web74. This project explored the application of the Open Annotation Model to the use case of capturing text mining output, by harmonizing the data models of the existing proposals.\n\nThe existing RDF-based representation of the PubAnnotation tool39 was used as a starting point, and adapted for compatibility with the Open Annotation model. The Open Annotation model provides an annotation class that relates a web resource to information that is about that resource; this representational choice is different from other models yet critically allows separation of metadata (e.g. provenance information) about the annotation itself, from meta-data about the content of the annotation75. Several core requirements for text-based annotations were identified: (1) representation of document spans as annotation targets; (2) representation of \"simple\" associations, e.g. between a span of text and a concept such as an ontology identifier; (3) representation of \"complex\" associations, e.g. between several spans of text and a relation or event. In addition, the overall structure of a document corpus, which can consist of several documents, must be modeled in such a way as to allow those documents to have internal structure such as chapters, sections, passages or sentences. PubAnnotation models text spans relative to these internal structural elements, while BioC and UIMA have adopted absolute character offsets across a complete document. The model developed here allows for both, by allowing the target of annotation to be either a full document, or a document element as appropriate. It is hoped that the proposals made for web-based document annotation representations will enable interoperability with other Open Annotation-based data and tools, while also addressing the need to move linguistic annotation into the web.\n\nDuring BioHackathon 2014, the integration of literature annotation resources was pursued with actual data sets. Colorado Richly Annotated Full-Text (CRAFT)76 is a recent important achievement of biomedical text mining, which included 67 full papers with rich annotation based on 7 biomedical ontologies.\n\nThe GRO corpus77 is a richly annotated corpus based on the Gene Regulation Ontology78. Allie is an acronym-annotated collection of all PubMed titles and abstracts79. They were all converted into PubAnnotation-compatible format, and submitted to PubAnnotation. The whole-PubMed-scale dataset, Allie, triggered the issue of scalability. However, integration of the two corpora, CRAFT and GRO, into PubAnnotation, demonstrated significantly improved utility.\n\nSPARQL is a standard language for querying triple stores. However, SPARQL queries can be difficult to write, even for experts. Usability studies have shown natural language interfaces to SPARQL to be the preferred method of SPARQL query formation assistance80. For this reason, software developers are encouraged to create applications that allow users to ask biomedical questions against triple stores using natural (i.e. human) language.\n\nBuilding on the work in BioHackathon 2012 on querying Systematized Nomenclature of Medicine-Clinical Terms (SNOMED-CT), during BioHackathon 2013 effort was focused on the Online Mendelian Inheritance in Man (OMIM) SPARQL endpoint, with a simultaneous focus on building an evaluation data set. Social networking was used to obtain use cases from biologists and informaticians, and it was quickly discovered that the system had an issue with differentiating between broad semantic types and specific instances. For example, “heart disease” was correctly mapped to a specific entity, but the word “genes” was incorrectly mapped to one specific gene. For this reason, dealing with the issue of recognizing broad semantic classes was the major focus of the development work, and testing semantic class recognition was the main focus of the testing effort. OMIM uses Type Unique Identifiers (TUIs), in the Unified Medical Language System (UMLS)81, to semantically type subjects and objects in its triple store, so we approached the problem of recognizing broad semantic classes as recognizing mentions of TUIs. Accordingly, a TUI concept recognizer was implemented into the open source LODQA system for automatic generation of SPARQL queries from natural language queries.\n\nEfforts to develop a natural language interface were continued in BioHackathon 2014, during which the LODQA system was configured for two large scale RDF datasets, Bio2RDF and BioGateway. In this way, it was demonstrated that technology like LODQA can answer a question like, “Which genes are involved in calcium binding?”, based on RDF data sets like Bio2RDF. However, it also revealed remaining performance issues.\n\n\nMetadata about RDF data resources\n\nBecause there is so far no solid guideline on publication of RDF data available, it is not clear for a researcher who wants to develop and release RDF data, how to create the associated metadata, how to describe the provenance of the data and how to assess the quality of the data/service. Also, understanding a dataset is not easy for users of data because there are so many classes, relations and possibilities. To resolve these issues, minimum requirements to represent statistics and characteristics of RDF data and services, including SPARQL endpoints, were discussed.\n\n\nDataset metadata\n\nThe International Society for Biocuration (ISB), in collaboration with the BioSharing forum, developed the BioDBCore82 which is a community-defined, uniform, generic description of the core attributes of biological databases. However, when it comes to the RDF datasets, one of the difficulties reported by users is that they find it difficult to figure out what data are in a dataset and how things are connected. Vocabulary of Interlinked Datasets (VoID) is a small vocabulary to describe key schemata style information about a dataset. It also includes key metadata such as when a dataset has been updated and under which license it falls. In this section, we propose a guideline for database providers, to provide useful extended VoID files for their users.\n\nAccess to consistent, high-quality metadata is critical to finding, understanding, and reusing scientific data - this is the core of the FAIR Data Principles83. However, while there are many relevant vocabularies for the annotation of a dataset, none sufficiently capture all the necessary metadata. Towards providing guidance for producing a high-quality description of biomedical datasets, we identified RDF vocabularies that could be used to specify common metadata elements and their value sets. The resulting guidelines, finalized under the auspices of the W3C Semantic Web for Health Care and the Life Sciences Interest Group (HCLSIG), cover elements of description, identification, versioning, attribution, provenance, and content summarization. This guideline reuses existing vocabularies, and is expected to meet key functional requirements including discovery, exchange, query, and retrieval.\n\nBig data presents an exciting opportunity to pursue large-scale analyses over collections of data in order to uncover valuable insights across a myriad of fields and disciplines. Yet, as more and more data are made available, researchers are finding it increasingly difficult to discover and reuse these data. One problem is that data are insufficiently described to understand what they are or how they were produced. A second issue is that no single vocabulary provides all key metadata fields required to support basic scientific use cases. For instance, the Data Catalog Vocabulary (DCAT) is used to describe datasets in catalogs, but does not deal with the issue of dataset evolution and versioning. A third issue is that data catalogs and data repositories all use different metadata standards, if they use any standard at all, and this prevents easy search, aggregation, and exchange of data descriptions. Thus, there is a need to combine these vocabularies in a comprehensive manner that meets the needs of data registries, data producers, and data consumers.\n\nWe developed a specification for the description of a dataset that meets key functional requirements (dataset description, linking, exchange, change, content summary), reuses 18 existing vocabularies, and is expressed using RDF. The specification covers 61 metadata elements pertaining to data description, identification, licensing, attribution, conformance, versioning, provenance, and content summary. Each metadata element includes a description and an example of use. The specification presents a three component model for modular description depending on whether specific files and versions are known (Figure 3). The summary level description focuses on release-independent information that mirrors the one captured by dataset registries; the distribution level description focuses on specific data files, their formats and downloadable location; and the version level description links summary descriptions with distribution descriptions. Each description level is bound to a different set of metadata requirements – mandatory, recommended, optional. A full worked example using the ChEMBL dataset is provided. The group is currently evaluating the specification with implementations for dataset registries such as Identifiers.org50 and IntegBio Database Catalog, as well as Linked Data repositories such as Bio2RDF84. The specification is available from the W3C site85.\n\nAs VoID is a vocabulary for describing datasets that can be used to generate documentation and assist users in finding key knowledge on how to write analytical data queries, the InterMine group worked on automatically generating VoID files for InterMine-based Model Organism Databases, while the UniProt group worked on the same for showing classes and predicates used in the named graphs on the UniProt SPARQL endpoint.\n\nInterMine86 is an open source graph-based data warehouse system built on top of PostgreSQL. Through a collaboration (the InterMOD consortium87), with most of the main animal Model Organism Databases (MODs) there are now InterMine databases available for budding yeast (SGD)88, rat (RGD89), zebrafish (ZFIN90), mouse (MGI91), nematode (WormBase, unpublished), Fly (InterMine group)92, and Arabidopsis93 with further MOD InterMine instances expected. Extensive data from the modENCODE project94 are also available through modMine95. As a step towards exposing these rich data as RDF, code was developed that uses existing InterMine RESTful web services to interrogate the FlyMine database and to generate a VoID description of the database. Further work is required to adjust the core InterMine data model to include additional database metadata items. This will then allow the automatic generation of VoID descriptions for any InterMine database. Further work is also required to ensure that appropriate standards are adhered to, especially for RDF predicates. In addition to the above developments, progress was made in creating a Sesame-based SPARQL endpoint for InterMine databases to complement the existing web application and web services. At the moment the endpoint only supports a small range of simple queries. It is hoped that in the future such endpoints will make available the rich data assembled and curated by the world wide Model Organism Database community. In the process this should provide opportunities for interoperation and also a mechanism for federation across the different resources.\n\nUniProt31 is available as RDF and can be queried via SPARQL and REST services. UniProt is a large and complicated database, that is difficult to explore due to its size. During the hackathon we implemented a procedure to generate a VoID file to describe UniProt data. The VoID file, now available on FTP and via the uniprot.org SPARQL Service Description (application/rdf+xml), is updated every release in synchrony with our production, and show users what types of data (and how much) are available in the UniProt datasets. We also document how many links to other databases UniProt provides, demonstrating the hub effect of UniProt.org in the life science domain. For the UniProt SPARQL endpoint this VoID description is used as a key part of the user documentation describing the schema of the UniProt data.\n\nSchema.org is a collection of extensible schemas that webmasters can use to mark up structured data on their web pages with the aim of improving search engine performance and enabling the creation of other applications. The initiative was founded by Google, Bing and Yahoo! as a collaboration to improve the web and their search results by using such structured data. More than 700 item types have been listed in schema.org, some of which have been supported by these search engines. If webmasters mark up their content in an acceptable markup format (e.g. Microdata, microformats or RDFa), then web crawler programs can detect these structured data and they can be rendered as rich snippets in the search results.\n\nDuring the BioHackathon, the members of this working group proposed two item types for a schema extension: \"BiologicalDatabase\" and \"BiologicalDatabaseEntry\". We discussed what item properties would be suitable for our purposes and how to label them in markup. Finally, we decided to use the Microdata format to mark up web pages and proposed five original properties: \"entryID\", \"isEntryOf\", “taxon”, \"seeAlso\" and \"reference\". Work in this area is now being carried forward by the bioschemas.org project.\n\nWe also publicized our proposal and encouraged BioHackathon members to mark up their databases. A Microdata crawler was created to extract these structured data. We modified \"Sagace\"96, a web-based search engine for biomedical data and resources in Japan, developed at the NIBIOHN in collaboration with NBDC. We confirmed that marked-up data showed up as rich snippets in search results. Ten databases have been marked up with our new proposal and so can help improve the readability of search results. This service is freely available at http://sagace.nibiohn.go.jp.\n\n\nProvenance of data\n\nSeveral models for associating provenance for an assertion have been proposed, but there has been inadequate evaluation to determine how accurately they are able to represent the myriad of provenance details required to support citation and reuse. The approach taken at BioHackathon 2014 was to survey and document assertional provenance methods, develop tools to populate these models, develop evaluation metrics to compare them, and assess this comparison. We describe a selection of these activities below.\n\nA nanopublication is defined as the smallest unit of publishable information that represents a finely-grained, but complete idea. Nanopublications are composed of such fine-grained assertions coupled with provenance metadata about the assertion, such as the methods used to create it and personal and institutional attributions, and finally additional metadata about the nanopublication itself, such as who or what created it, and when. The aim is to make a formal, predictable, and transparent relationship between data and its provenance. Nanopublications will be discussed here with respect to their application to FANTOM597,98 data, to track DBCLS literature curation, and within the Semantic Automated Discovery and Integration (SADI) framework99.\n\nThe FANTOM5 project monitored transcription initiation at single base-pair resolution in mammalian genomes by Cap Analysis Gene Expression (CAGE) coupled with single molecule sequencing97,98. Promoters were defined as upstream of CAGE peaks (transcription start site clusters) and their activities were quantified based on their read counts. The FANTOM5 promoters and their activities were described in nanopublications100 to facilitate their open and interoperable exchange. Three classes of nanopublications, having the following assertions101, were generated: 1) A CAGE peak is defined in a specific region of the genome, 2) The CAGE peak is a transcription start site (TSS) region, which is part of a gene, 3) The CAGE peak is active at a certain level in a specified sample. Class 1 nanopublications (CAGE peaks) provide minimum information based on a model on genomic coordinates. They can be exported to genome browsers. Class 2 nanopublications (gene associations) are served as supplemental data to allow biological searches. This class of nanopublications may be re-released when a new data processing workflow is available or when different parameters or gene definitions are used. Class 3 nanopublications (activity levels of transcription in individual samples) are used only if the details of expression are relevant in a given biological search. By dissecting the whole data set into three classes of nanopublications with different granularities, its reusability is increased. These nanopublications are available at http://rdf.biosemantics.org, and they have been reported also in an article related to FANTOM5101.\n\nThe DBCLS has developed a web-based gene annotation tool, TogoAnnotation, has provides an easy way of accessing and adding annotations. Likewise Gene Indexing was developed as a simple named-entity recognition (NER) task in order to make connections between genomic loci and the literature. Gene Indexing generates micro-annotations by manually extracting gene and protein symbols from the text, tables and figures of full papers and connecting them to both PubMed IDs and genome location. A total of 10 curators cooperated over a five year period to manually annotate over 5,000 full papers relating to microbes. In this way over 200,000 gene/protein micro-annotations were generated.\n\nBased on the above data, during the BioHackathon 2014, a Nanopublication model was developed for these literature curation data, as well as a converter to make any annotation in the TogoAnnotation system representable as a Nanopublication RDF (Figure 4). It is intended that the curation data be integrated into the TogoGenome system and be expanded as a standard distributed annotation platform in the future.\n\nThe SADI Semantic Web services project also has a need to represent rich provenance data regarding how its services create their output. Given the rapid growth and notable success of the OpenPHACTS102 and NanoPublications103 projects, it seems desirable that analytical Services - those following the SADI Semantic Web Service design patterns in particular - should output semantic data that follows the same NanoPublication paradigm. This would allow SADI services to publish new biomedical knowledge directly into the vast integrated NanoPublications space, and take advantage of their integration tools.\n\nExtensions to the existing Perl SADI::Simple codebase in Comprehensive Perl Archive Network (CPAN) were undertaken at the hackathon. A key consideration was to ensure that the code could support distinct metadata for each triple, since SADI services are specifically designed to support multiplexed inputs potentially spread over a large number of processors for analysis, before being reassembled into an output message. As such, it is potentially the case that each triple has slightly distinct provenance information. The implemented solution guarantees globally unique identification of each of these nanopublications, for each execution, even over multiple iterations of the same input data.\n\nNanoPublications are created when, through HTTP content negotiation, the client requests n-quads. The service responds with an RDF structure that follows the structure of the (proposed) NanoPublication Collection.\n\nRequesting quads from a ‘legacy’ SADI Service that does not support NanoPublications will result in a HTTP 406 (Not Acceptable) response, with an output body in application/rdf-xml, as is allowed by HTTP 1.1.\n\nDiscovering and reusing data requires substantial expertise about where data are located and how to transform them into a more useable form for further analysis. While the Bio2RDF project transforms dozens of key bioinformatic resources into RDF, and is made available through public SPARQL endpoints, a key challenge still remained: how to identify which datasets contain the entities and relations that are of interest to solve a particular problem. To this end, Bio2RDF now generates and publishes summaries of the dataset contents in each of its SPARQL endpoints, thereby simplifying lookup, and reducing server load for expensive and common queries.\n\nDuring the hackathon, an architecture was developed for an automated approach that utilizes the metadata from Bio2RDF’s content summaries to automatically generate SADI Semantic Web Services that provide discoverable access to this Bio2RDF data104. SADI Services use ontologies to formally describe their inputs and outputs, such that it is possible to find services of interest by querying their ontological descriptions via a global Service metadata registry. In the case of these Bio2RDF SADI Services, the input data-type is a simple Bio2RDF typed-URI (for example [http://bio2rdf.org/mesh:C025643 rdf:type ctd:Chemical]) and the output is, as per the SADI specifications, the input node annotated with a Bio2RDF relation (for example [http://bio2rdf.org/mesh:C025643 sio:is-participant-in http://bio2rdf.org/go:0008380). Such metadata descriptions can be automatically generated from the Bio2RDF indexes, and moreover, the corresponding SPARQL queries that make up the business logic of the service can similarly be automatically constructed based on the information in these indexes. As such, both the service description, as well as the service itself, can be dynamically created to provide access to any Bio2RDF data of interest.\n\nThe advantage of exposing Bio2RDF as a set of SADI services is that the data in Bio2RDF becomes discoverable - software does not need to know, a priori, what data/relations exist in which Bio2RDF endpoint. Moreover, when exposed as SADI Services, Bio2RDF data can more easily be integrated into workflows using popular workflow editors such as Taverna105 or as demonstrated by our use of these services within Galaxy workflows106.\n\nA large amount of biomedical information is available via SPARQL endpoints, often in a redundant way. Life Sciences databases often integrate information from different sources to enrich the data they provide, and some information resources are pure aggregators whose value is in the harmonization of the information that they collect. As these resources publish their information on the Semantic Web, the result is that the same information is present in multiple endpoints. As a consequence, to decide which endpoint to use to access some particular data of interest is not a trivial task. Two hackathon activities addressed this issue. The development of a dataset descriptor is useful to know which data are present in an endpoint, with information on version, representation and update policies. But even if such a descriptor is provided, there is still an issue of the reliability of endpoints. It is also difficult to know which endpoints are actively maintained and which are not.\n\nYummyData is a project that monitors endpoints by periodically running queries and performing a few tests. By collecting data over extended periods, it can provide a proxy for the reliability of an endpoint and the dynamism of the information it provides. More specifically, YummyData periodically queries datahub.io for datasets tagged as being of biomedical interest. It combines the result with a list of curated endpoints and, periodically, it runs a series of tests and queries and stores their results. YummyData performs some tests to determine whether the endpoint provides a VoID descriptor (see section above), as well as to measure response time. It also runs a series of queries that can be generic or endpoint-specific. Generic queries inspect aggregate information such as the number of statements, distinct resources, or properties. Specific queries are currently only implemented as a proof of concept, but they are intended to reveal aspects of the quality of the data provided by endpoints. For instance, a typical query would ask for the number of entities annotated via a given evidence code. Results over time are then aggregated in two types of rating: a SPARQL score that is a numeric value that results from a count of positive response codes over time windows; a star rating that is intended to provide a more qualitative assessment of features (e.g. the availability of a valid VoID descriptor, or of a copyright notice, yields +1 star). At the time of writing, YummyData has collected data for about a year on a few tens of information resources. A subset of these data are accessible via the http://yummydata.org website.\n\n\nConclusion\n\nTo fulfil the mission of the DBCLS, which is to integrate life sciences databases, the annual BioHackathon series was started in 2008 to explore state-of-the-art technological solutions. The utilization of Semantic Web technologies as a means for database integration was introduced in BioHackathon 20103. Since then, we have collaboratively worked as a community to promote the use of RDF and ontologies in life sciences. As one of the demonstration products, DBCLS released the first RDF-based genome database, TogoGenome, in 2013. Subsequently, the EBI RDF Platform was released by EMBL-EBI and PubChem RDF was published by NCBI, and these provide fundamental database resources in genomics to the wider biomedical research community as well as the pharmaceutical and biotechnology industries. The NBDC RDF portal launched in 2015 complements the above resources by adding other major domains such as protein structures and glycoscience resources. The 6th and 7th BioHackathons in 2013 and 2014 were held to develop and improve methods and best practices for creating and publishing these community wide resources. As a result, the field is becoming ready for testing in real world use cases such as dealing with human genome-scale biomedical data. Other domains (e.g. plants/crops) are less developed but gaining momentum (see for instance AgroPortal). At the same time we found another layer of demands for additional development in real world applications such as genotype-phenotype information to drug discovery, which define further challenges and will be addressed in the upcoming BioHackathons.\n\n\nData availability\n\nNo data are associated with this article.\n\nRecords of the BioHackathon 2013 and 2014 meetings are aggregated at https://github.com/dbcls/bh13/wiki and https://github.com/dbcls/bh14/wiki respectively.\n\nZenodo: dbcls/bh13: Included repositories related to BH13. http://doi.org/10.5281/zenodo.3271508107\n\nThis project contains the following extended data:\n\ndbcls/bh13-v1.0.0.zip (BioHackathon 2013 records)\n\nZenodo: dbcls/bh14: Included repositories related to BH14. http://doi.org/10.5281/zenodo.3271509108\n\nThis project contains the following extended data:\n\ndbcls/bh14-v1.0.0.zip (BioHackathon 2014 records)\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).", "appendix": "Grant information\n\nFunding for the BioHackathon was supported by the National Bioscience Database Center (NBDC) of the Japan Science and Technology Agency (JST).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nBioHackathon 2013 and 2014 are supported by the Integrated Database Project (Ministry of Education, Culture, Sports, Science and Technology of Japan) and hosted by the National Bioscience Database Center (NBDC) and the Database Center for Life Science (DBCLS). We thank Yuji Kohara, the director of DBCLS, for his supporting the BioHackathons.\n\n\nReferences\n\nKatayama T, Arakawa K, Nakao M, et al.: The DBCLS BioHackathon: standardization and interoperability for bioinformatics web services and workflows. The DBCLS BioHackathon Consortium. J Biomed Semantics. 2010; 1(1): 8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKatayama T, Wilkinson MD, Vos R, et al.: The 2nd DBCLS BioHackathon: interoperable bioinformatics Web services for integrated applications. J Biomed Semantics. 2011; 2: 4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKatayama T, Wilkinson MD, Micklem G, et al.: The 3rd DBCLS BioHackathon: improving life science data integration with Semantic Web technologies. J Biomed Semantics. 2013; 4(1): 6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKatayama T, Wilkinson MD, Aoki-Kinoshita KF, et al.: BioHackathon series in 2011 and 2012: penetration of ontology and linked data in life science domains. J Biomed Semantics. 2014; 5(1): 5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJupp S, Malone J, Bolleman J, et al.: The EBI RDF platform: linked open data for the life sciences. Bioinformatics. 2014; 30(9): 1338–1339. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMagrane M, UniProt Consortium: UniProt Knowledgebase: a hub of integrated protein data. Database (Oxford). 2011; 2011: bar009. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBolton EE, Wang Y, Thiessen PA, et al.: PubChem: Integrated Platform of Small Molecules and Biological Activities. Annu Rep Comput Chem. 2008; 4: 217–241. Publisher Full Text\n\nUchiyama I, Mihara M, Nishide H, et al.: MBGD update 2015: microbial genome database for flexible ortholog analysis utilizing a diverse set of genomic data. Nucleic Acids Res. 2015; 43(Database issue): D270–D276. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKatayama T, Kawashima S, Okamoto S, et al.: TogoGenome/TogoStanza: modularized Semantic Web genome database. Database (Oxford). 2019; 2019: bay132. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBolleman JT, Mungall CJ, Strozzi F, et al.: FALDO: a semantic standard for describing the location of nucleotide and protein feature annotation. J Biomed Semantics. 2016; 7: 39. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBaran J, Durgahee BS, Eilbeck K, et al.: GFVO: the Genomic Feature and Variation Ontology. PeerJ. 2015; 3: e933. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCochrane G, Karsch-Mizrachi I, Takagi T, et al.: The international nucleotide sequence database collaboration. Nucleic Acids Res. 2016; 44(D1): D48–D50. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAken BL, Achuthan P, Akanni W, et al.: Ensembl 2017. Nucleic Acids Res. 2017; 45(D1): D635–D642. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLandrum MJ, Lee JM, Benson M, et al.: ClinVar: public archive of interpretations of clinically relevant variants. Nucleic Acids Res. 2016; 44(D1): D862–D868. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHamosh A, Scott AF, Amberger JS, et al.: Online Mendelian Inheritance in Man (OMIM), a knowledgebase of human genes and genetic disorders. Nucleic Acids Res. 2005; 33(Database issue): D514–517. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFokkema IF, Taschner PE, Schaafsma GC, et al.: LOVD v.2.0: the next generation in gene variant databases. Hum Mutat. 2011; 32(5): 557–563. PubMed Abstract \| Publisher Full Text\n\nStenson PD, Mort M, Ball EV, et al.: The Human Gene Mutation Database: building a comprehensive mutation repository for clinical and molecular genetics, diagnostic testing and personalized genomic medicine. Hum Genet. 2014; 133(1): 1–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nImanishi T, Itoh T, Suzuki Y, et al.: Integrative annotation of 21,037 human genes validated by full-length cDNA clones. PLoS Biol. 2004; 2(6): e162. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTakeda JI, Yamasaki C, Murakami K, et al.: H-InvDB in 2013: an omics study platform for human functional gene and transcript discovery. Nucleic Acids Res. 2013; 41(Database issue): D915–D919. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBurge SW, Daub J, Eberhardt R, et al.: Rfam 11.0: 10 years of RNA families. Nucleic Acids Res. 2013; 41(Database issue): D226–32. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKozomara A, Griffiths-Jones S: miRBase: integrating microRNA annotation and deep-sequencing data. Nucleic Acids Res. 2011; 39(Database issue): D152–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLewis BP, Burge CB, Bartel DP: Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets. Cell. 2005; 120(1): 15–20. PubMed Abstract \| Publisher Full Text\n\nO’Leary NA, Wright MW, Brister JR, et al.: Reference sequence (RefSeq) database at NCBI: current status, taxonomic expansion, and functional annotation. Nucleic Acids Res. 2016; 44(D1): D733–D745. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRosse C, Mejino JLV Jr: The Foundational Model of Anatomy Ontology. In Anat Ontol Bioinforma Princ Pract. Computational Biol. Edited by Burger A, Davidson D, Baldock R. Springer; 2008; 6: 59–117. Publisher Full Text\n\nBard JB: The AEO, an Ontology of Anatomical Entities for Classifying Animal Tissues and Organs. Front Genet. 2012; 3(FEB): 18. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDahdul WM, Balhoff JP, Blackburn DC, et al.: A unified anatomy ontology of the vertebrate skeletal system. PLoS One. 2012; 7(12): e51070. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRamírez MJ, Coddington JA, Maddison WP, et al.: Linking of digital images to phylogenetic data matrices using a morphological ontology. Syst Biol. 2007; 56(2): 283–294. PubMed Abstract \| Publisher Full Text\n\nChiba H, Nishide H, Uchiyama I: Construction of an ortholog database using the semantic web technology for integrative analysis of genomic data. PLoS One. 2015; 10(4): e0122802. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMiñarro-Gimenez JA, Madrid M, Fernandez-Breis JT: OGO: an ontological approach for integrating knowledge about orthology. BMC Bioinformatics. 2009; 10(Suppl 10): S13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchmitt T, Messina DN, Schreiber F, et al.: Letter to the editor: SeqXML and OrthoXML: standards for sequence and orthology information. Brief Bioinform. 2011; 12(5): 485–8. PubMed Abstract \| Publisher Full Text\n\nUniProt Consortium: UniProt: A hub for protein information. Nucleic Acids Res. 2015; 43(Database issue): D204–D212. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHakenberg J, Plake C, Leaman R, et al.: Inter-species normalization of gene mentions with GNAT. Bioinformatics. 2008; 24(16): i126–132. PubMed Abstract \| Publisher Full Text\n\nWei CH, Kao HY: Cross-species gene normalization by species inference. BMC Bioinformatics. 2011; 12(Suppl 8): S5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWei CH, Harris BR, Kao HY, et al.: tmVar: a text mining approach for extracting sequence variants in biomedical literature. Bioinformatics. 2013; 29(11): 1433–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCaporaso JG, Baumgartner WA Jr, Randolph DA, et al.: MutationFinder: a high-performance system for extracting point mutation mentions from text. Bioinformatics. 2007; 23(14): 1862–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeaman R, Gonzalez G: BANNER: an executable survey of advances in biomedical named entity recognition. Pac Symp Biocomput. 2008; 13: 652–663. PubMed Abstract \| Publisher Full Text\n\nGerner M, Sarafraz F, Bergman CM, et al.: BioContext: an integrated text mining system for large-scale extraction and contextualization of biomolecular events. Bioinformatics. 2012; 28(16): 2154–61. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVan Landeghem S, Björne J, Wei CH, et al.: Large-scale event extraction from literature with multi-level gene normalization. PLoS One. 2013; 8(4): e55814. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKim J, Wang Y: PubAnnotation - a persistent and sharable corpus and annotation repository. In Proc 2012 Work Biomed Nat Lang Process. 2012(BioNLP): 202–205. Reference Source\n\nLi H, Handsaker B, Wysoker A, et al.: The Sequence Alignment/Map format and SAMtools. Bioinformatics. 2009; 25(16): 2078–2079. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDanecek P, Auton A, Abecasis G, et al.: The variant call format and VCFtools. Bioinformatics. 2011; 27(15): 2156–2158. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nReese MG, Moore B, Batchelor C, et al.: A standard variation file format for human genome sequences. Genome Biol. 2010; 11(8): R88. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFernández JD, Martínez-Prieto MA, Gutiérrez C, et al.: Binary RDF representation for publication and exchange (HDT). J Web Semant. 2013; 19: 22–41. Publisher Full Text\n\nFujisawa T, Narikawa R, Maeda SI, et al.: CyanoBase: a large-scale update on its 20th anniversary. Nucleic Acids Res. 2017; 45(D1): D551–D554. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBuels R, Yao E, Diesh CM, et al.: JBrowse: a dynamic web platform for genome visualization and analysis. Genome Biol. 2016; 17: 66. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKalderimis A, Lyne R, Butano D, et al.: InterMine: extensive web services for modern biology. Nucleic Acids Res. 2014; 42(Web Server issue): W468–472. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVelankar S, Dana JM, Jacobsen J, et al.: SIFTS: Structure Integration with Function, Taxonomy and Sequences resource. Nucleic Acids Res. 2013; 41(Database issue): D483–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKinjo AR, Bekker GJ, Suzuki H, et al.: Protein Data Bank Japan (PDBj): updated user interfaces, resource description framework, analysis tools for large structures. Nucleic Acids Res. 2017; 45(D1): 282–288. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIson J, Kalas M, Jonassen I, et al.: EDAM: an ontology of bioinformatics operations, types of data and identifiers, topics and formats. Bioinformatics. 2013; 29(10): 1325–32. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJuty N, Le Novère N, Laibe C: Identifiers.org and MIRIAM Registry: community resources to provide persistent identification. Nucleic Acids Res. 2012; 40(Database issue): D580–586. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKomiyama Y, Masaki B, Saad G, et al.: UTProt: Database Integration and Tool Development for Intractomics Utilizing Biosemantics. In 2013 Annu Conv Japanese Soc Bioinforma. 2013; 81–82.\n\nPerez-Riverol Y, Hermjakob H, Kohlbacher O, et al.: Computational proteomics pitfalls and challenges: HavanaBioinfo 2012 Workshop report. J Proteomics. 2013; 87: 134–138. PubMed Abstract \| Publisher Full Text\n\nDeutsch EW, Albar JP, Binz PA, et al.: Development of data representation standards by the human proteome organization proteomics standards initiative. J Am Med Informatics Assoc. 2015; 22(3): 495–506. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPerez-Riverol Y, Alpi E, Wang R, et al.: Making proteomics data accessible and reusable: current state of proteomics databases and repositories. Proteomics. 2015; 15(5–6): 930–49. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPerez-Riverol Y, Wang R, Hermjakob H, et al.: Open source libraries and frameworks for mass spectrometry based proteomics: a developer's perspective. Biochim Biophys Acta. 2014; 1844(1 Pt A): 63–76. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChervitz SA, Deutsch EW, Field D, et al.: Data standards for Omics data: the basis of data sharing and reuse. In Methods Mol Biol. Edited by Mayer B. Springer; 2011; 719: 31–69. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGriss J, Jones AR, Sachsenberg T, et al.: The mzTab data exchange format: communicating mass-spectrometry-based proteomics and metabolomics experimental results to a wider audience. Mol Cell Proteomics. 2014; 13(10): 2765–2775. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSugimoto M, Kawakami M, Robert M, et al.: Bioinformatics Tools for Mass Spectroscopy-Based Metabolomic Data Processing and Analysis. Curr Bioinform. 2012; 7(1): 96–108. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPerez-Riverol Y, Uszkoreit J, Sanchez A, et al.: ms-data-core-api: an open-source, metadata-oriented library for computational proteomics. Bioinformatics. 2015; 31(17): 2903–2905. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCraig R, Cortens JP, Beavis RC: Open source system for analyzing, validating, and storing protein identification data. J Proteome Res. 2004; 3(6): 1234–1242. PubMed Abstract \| Publisher Full Text\n\nVizcaíno JA, Côté RG, Csordas A, et al.: The Proteomics Identifications (PRIDE) database and associated tools: Status in 2013. Nucleic Acids Res. 2013; 41(Database issue): D1063–1069. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAoki-Kinoshita KF, Bolleman J, Campbell MP, et al.: Introducing glycomics data into the Semantic Web. J Biomed Semantics. 2013; 4(1): 39. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRanzinger R, Aoki-Kinoshita KF, Campbell MP, et al.: GlycoRDF: an ontology to standardize glycomics data in RDF. Bioinformatics. 2015; 31(6): 919–925. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAshburner M, Ball CA, Blake JA, et al.: Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet. 2000; 25(1): 25–29. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDegtyarenko K, de matos P, Ennis M, et al.: ChEBI: a database and ontology for chemical entities of biological interest. Nucleic Acids Res. 2008; 36(Database issue): D344–350. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcDonald AG, Boyce S, Tipton KF: ExplorEnz: the primary source of the IUBMB enzyme list. Nucleic Acids Res. 2009; 37(Database issue): D593–597. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKotera M, Nishimura Y, Nakagawa Z, et al.: PIERO ontology for analysis of biochemical transformations: effective implementation of reaction information in the IUBMB enzyme list. J Bioinform Comput Biol. 2014; 12(6): 1442001. PubMed Abstract \| Publisher Full Text\n\nYamanishi Y, Tabei Y, Kotera M: Metabolome-scale de novo pathway reconstruction using regioisomer-sensitive graph alignments. Bioinformatics. 2015; 31(12): i161–i170. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTabei Y, Yamanishi Y, Kotera M: Simultaneous prediction of enzyme orthologs from chemical transformation patterns for de novo metabolic pathway reconstruction. Bioinformatics. 2016; 32(12): i278–i287. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIde N, Suderman K: GrAF: A Graph-based Format for Linguistic Annotations. In Proc Linguist Annot Work 2007. 2007; 1–8. Reference Source\n\nComeau DC, Islamaj Doğan R, Ciccarese P, et al.: BioC: a minimalist approach to interoperability for biomedical text processing. Database (Oxford). 2013; 2013: bat064. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCiccarese P, Ocana M, Garcia Castro LJ, et al.: An open annotation ontology for science on web 3.0. J Biomed Semantics. 2011; 2 Suppl 2: S4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFerrucci D, Lally A: UIMA: an architectural approach to unstructured information processing in the corporate research environment. J Nat Lang Enginnering. 2004; 10: 327–348. Publisher Full Text\n\nSanderson R, Ciccarese P, Van de Sompel H: Designing the W3C open annotation data model. Proceedings of the 5th Annual ACM Web Science Conference-WebSci ’13. 2013; 366–375. Publisher Full Text\n\nVerspoor K, Ave E: Towards Adaptation of Linguistic Annotations to Scholarly Annotation Formalisms on the Semantic Web. In Proc Sixth Linguist Annot Work.2012; 75–84. Reference Source\n\nBada M, Eckert M, Evans D, et al.: Concept annotation in the CRAFT corpus. BMC Bioinformatics. 2012; 13: 161. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKim J, Han X, Lee V, et al.: GRO Task: Populating the Gene Regulation Ontology with events and relations. Work BioNLP Shar Task. 2013; 50–57. Reference Source\n\nBeisswanger E, Lee V, Kim JJ, et al.: Gene Regulation Ontology (GRO): design principles and use cases. Stud Health Technol Inform. 2008; 136: 9–14. PubMed Abstract\n\nYamamoto Y, Yamaguchi A, Bono H, et al.: Allie: a database and a search service of abbreviations and long forms. Database (Oxford). 2011; 2011:bar013. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKaufmann E, Bernstein A: How useful are natural language interfaces to the semantic Web for casual end-users? Lect Notes Comput Sci (including Subser Lect Notes Artif Intell Lect Notes Bioinformatics). 2007; 4825 LNCS: 281–294. Publisher Full Text\n\nBodenreider O: The Unified Medical Language System (UMLS): integrating biomedical terminology. Nucleic Acids Res. 2004; 32(Database issue): D267–D270. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGaudet P, Bairoch A, Field D, et al.: Towards BioDBcore: A community-defined information specification for biological databases. Database (Oxford). 2011; 2011: baq027. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWilkinson MD, Dumontier M, Aalbersberg IJ, et al.: The FAIR Guiding Principles for scientific data management and stewardship. Sci Data. 2016; 3: 160018. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCallahan A, Cruz-Toledo J, Dumontier M: Ontology-Based Querying with Bio2RDF’s Linked Open Data. J Biomed Semantics. 2013; 4 Suppl 1: S1. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDumontier M, Gray AJG, Marshall MS, et al.: The health care and life sciences community profile for dataset descriptions. PeerJ. 2016; 4: e2331. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSmith RN, Aleksic J, Butano D, et al.: InterMine: a flexible data warehouse system for the integration and analysis of heterogeneous biological data. Bioinformatics. 2012; 28(23): 3163–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSullivan J, Karra K, Moxon SA, et al.: InterMOD: integrated data and tools for the unification of model organism research. Sci Rep. 2013; 3: 1802. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBalakrishnan R, Park J, Karra K, et al.: YeastMine--an integrated data warehouse for Saccharomyces cerevisiae data as a multipurpose tool-kit. Database (Oxford). 2012; 2012: bar062. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang SJ, Laulederkind SJ, Hayman GT, et al.: Analysis of disease-associated objects at the Rat Genome Database. Database (Oxford). 2013; 2013: bat046. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHowe DG, Bradford YM, Conlin T, et al.: ZFIN, the Zebrafish Model Organism Database: increased support for mutants and transgenics. Nucleic Acids Res. 2013; 41(Database issue): D854–60. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMotenko H, Neuhauser SB, O’Keefe M, et al.: MouseMine: a new data warehouse for MGI. Mamm Genome. 2015; 26(7–8): 325–330. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLyne R, Smith R, Rutherford K, et al.: FlyMine: an integrated database for Drosophila and Anopheles genomics. Genome Biol. 2007; 8(7): R129. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKrishnakumar V, Contrino S, Cheng CY, et al.: ThaleMine: A Warehouse for Arabidopsis Data Integration and Discovery. Plant Cell Physiol. 2017; 58(1): e4. PubMed Abstract \| Publisher Full Text\n\nCelniker SE, Dillon LA, Gerstein MB, et al.: Unlocking the secrets of the genome. Nature. 2009; 459(7249): 927–930. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nContrino S, Smith RN, Butano D, et al.: modMine: flexible access to modENCODE data. Nucleic Acids Res. 2012; 40(Database issue): D1082–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMorita M, Igarashi Y, Ito M, et al.: Sagace: a web-based search engine for biomedical databases in Japan. BMC Res Notes. 2012; 5: 604. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nForrest ARR, Kawaji H, Rehli M, et al.: A promoter-level mammalian expression atlas. Nature. 2014; 507(7493): 462–470. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAndersson R, Gebhard C, Miguel-Escalada I, et al.: An atlas of active enhancers across human cell types and tissues. Nature. 2014; 507(7493): 455–461. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWilkinson MD, Vandervalk B, McCarthy L: The Semantic Automated Discovery and Integration (SADI) Web service Design-Pattern, API and Reference Implementation. J Biomed Semantics. 2011; 2(1): 8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMons B, Velterop J: Nano-Publication in the e-science era. In Proc Work Semant Web Appl Sci Discourse (SWASD 2009). 2009. Reference Source\n\nLizio M, Harshbarger J, Shimoji H, et al.: Gateways to the FANTOM5 promoter level mammalian expression atlas. Genome Biol. 2015; 16: 22. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHarland L: Open PHACTS: A Semantic Knowledge Infrastructure for Public and Commercial Drug Discovery Research. In Knowl Eng Knowl Manag EKWQ 2012 Lect Notes Comput Sci. Springer Berlin Heidelberg; 2012; 7603: 1–7. Publisher Full Text\n\nKuhn T, Barbano PE, Nagy ML, et al.: Broadening the Scope of Nanopublications. In Semant Web Semant Big Data. Edited by Cimianoe P, Corcho O, Presutti V, Hollink L, Rudolph S. Springer Berlin Heidelberg; 2013; 7882: 487–501. Publisher Full Text\n\nGonzález AR, Callahan A, Cruz-toledo J, et al.: Automatically exposing OpenLifeData via SADI semantic Web Services. J Biomed Semantics. 2014; 5: 46. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWolstencroft K, Haines R, Fellows D, et al.: The Taverna workflow suite: designing and executing workflows of Web Services on the desktop, web or in the cloud. Nucleic Acids Res. 2013; 41(Web Server issue): W557–561. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAranguren ME, Wilkinson MD: Enhanced reproducibility of SADI web service workflows with Galaxy and Docker. Gigascience. 2015; 4(1): 59. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKatayama T: dbcls/bh13: Added CC-BY license as requested by the journal (Version 1.0.1). Zenodo. 2019. http://www.doi.org/10.5281/zenodo.3271508\n\nKatayama T: dbcls/bh14: Added CC-BY license as requested by the journal (Version 1.0.1). Zenodo. 2019. http://www.doi.org/10.5281/zenodo.3271509" }	[ { "id": "54207", "date": "18 Nov 2019", "name": "Todd J. Vision", "expertise": [ "Reviewer Expertise Evolutionary genetics", "genome evolution", "bioinformatics and computational biology." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors report on the many activities conducted under the umbrella of the BioHackathons 2013 and 2014. While there are recurring intellectual threads, most notably a focus on support for RDF, the manuscript is really a collection of reports of the progress made by each of the different subgroups involved in these events. I admire the work that must have gone into fitting each of these individual efforts into a coherent narrative, and this is a very useful historical document of those efforts.\nI do have some suggestions below where I think the manuscript could be improved - primarily to help developers and users put this work into context. Most of these are are not critical flaws, however, and I would leave to the authors' discretion how they wish to respond to these comments and suggestions.\n\nThe one issue I think is critical for acceptance is the explanation of Figure 1 (see below).\nOne recurring issue is the ambiguity of the reference time frame, both for the starting and stopping points of the work described. This is important since the report covers work that took place 5-6 years in the past, a long time ago in the world of bioinformatics.\n\nWith respect to the starting point, the use of present tense makes it unclear in many passages whether the authors are describing the state of software as of 2013/2014, or today. To avoid confusion, I would recommend either fixing the relevant passages throughout (I list some of them below) or clarify upfront that the state of affairs of the software tools and resources described is as of 5-6 years ago, which in many cases will no longer be accurate.\nRegarding the stopping point, it is not clear if the progress described reflects what was accomplished during the event, up until the next biohackathon, or some other time point.\nThis is perhaps a larger ask, but I wanted as a reader to know where these efforts stand today. Putting the work in context of the progress since would take a weakness of the manuscript (ie. reporting on older work) and turn it into a virtue (evaluating the impact each of these efforts had on their fields with the benefits of a few years of hindsight). Perhaps this could be summarized by a \"where-are-they-now\" table.\nIn addition, there are a number of claims made, particularly regarding tool performance, without supporting evidence provided. In some cases, these might be from subjective assessments, which is fine as long as that is clear. But subjective or not, the lack of evidence makes the claims difficult to evaluate. I point some of these passages out below.\nCOMMENTS BY SECTION:\nABSTRACT\n\"an exciting and unanticipatable variety of real world applications in coming years\" - this is an odd statement given that several years have already passed.\nINTRODUCTION\n\"To fully utilize open data in life sciences, semantic interoperability and standardization of data are required to allow innovative development of applications\" - while it may be true, this claim is not directly supported. A weaker claim might be more justified.\nIMPROVEMENT AND UTILIZATION OF RDF DATA IN LIFE SCIENCES\" \"RDF ... is becoming widely accepted\" - can the other cite support for this claim?\nGENOMIC INFORMATION\nFIGURE 1 - needs further explanation via a caption and/or legend. What do the colors represent, and the different types of lines? The use of URLs deserves explanation - especially the one to a google doc.\nIDENTIFIERS FOR SEQUENCES AND ANNOTATIONS\nI couldn't quite figure out what I was supposed to understand about the conceptual differences among the identifier models.\n\nTOOLS FOR SEMANTIC GENOME DATA\n\"To support these efforts we have collaborated with DDBJ, UniProt, and the EBI RDF group to develop ontologies for representing locations and annotations of genome sequences and used these developments for all prokaryotic genomes and, later, eukaryotic genomes\" - this passage seems redundant with the descriptions above. Also, the use of \"we\" here and in next sentence appears to be legacy text referring to only the TogoGenome participants.\nJBrowse on RDF was comparable in performance to typical RDB back-end - it would be good if the authors could support that claim with some of the data obtained.\nThe sections \"Visualization of semantic annotations in JBrowse\" and \"TogoGenome\" are somewhat redundant with each other. I recommend merging.\nINTERMINE \"An area that needs work, and is receiving attention,\" - is this still true 5-6 years later?\n\"Implement a simple adaptor allowing, as described above, JBrowse to request data directly from InterMine RESTful web services\" If this was actually described above, I missed it.\nTEXT-MINING\nI would personally find a screenshot of the text annotations in JBrowse much more interesting than the current Figure 2, which is exactly what you'd expect it to look like and doesn't provide evidence for the performance claim above.\nPROTEIN STRUCTURES AND EXPRESSIONS\n\"the Worldwide Protein Data Bank (wwPDB) will abolish the conventional PDB format\" and \"wwPDB will start assigning protein chain identifiers, which will also be encoded as URIs in the wwPDB/RDF.\" 2016 is three years ago, so if this has already been accomplished, this should be in past tense.\n\"Recently, the mzTab data exchange format was introduced\" - recently meaning 2014?\n\"the \"omics\" group in the BioHackathon 2014 made the first steps towards the development of the ProteomeXchange Interface (PROXI) for protein expression data exchange59.\" - my impression is that this effort has evolved considerably since 2014. If true, it would be valuable to put this into context for how the technology is used currently by the ProteomeXchange consortium.\n\n\"integration of the two corpora, CRAFT and GRO, into PubAnnotation, demonstrated significantly improved utility\" - is there supporting evidence for this claim?\n\"Access to consistent, high-quality metadata is critical to finding, understanding, and reusing scientific data - this is the core of the FAIR Data Principles83.\" - it's odd to reference this manifesto of principles that postdates the hackathons by at least two years. This would be less awkward if there was a table or section devoted to developments since the 2014, i.e. \"where are they now?\"\nDATASET METADATA\n\"The resulting guidelines, finalized under the auspices of the W3C Semantic Web for Health Care and the Life Sciences Interest Group (HCLSIG)\" - there's no link or citation given here.\n\nThe 4th pgph redundantly includes the same coverage list: \"..elements of description, identification, versioning, attribution, provenance, and content summarization.\"\n\"The group is currently evaluating the specification with implementations for dataset registries such as Identifiers.org50 and IntegBio Database Catalog, as well as Linked Data repositories such as Bio2RDF84.\" - update on progress since 2014?\nW2C link appears to be dead, but a paper was published in 2016. I get the impression that there was further work on this post-hackathon not mentioned here.\nVOID FOR INTERMINE AND UNIPROT\nThe relationship of prior work on VOiD (if any) to the outcome from the hackathon is not clear to me.\n\"there are now InterMine databases available\" - is \"now\" as of writing, or as of the time of the hackathon?\nThe VoID file [for Uniprot] is updated every release in synchrony with our production\" - replace use of \"our\" which I presume refers to the Uniprot team and not to the collaborative team of authors here.\nSCHEMA>ORG AND RDFA FOR BIOLOGICAL DATABASES\nwhat's \"NIBIOHN\"?\nNANOPUBLICATIONS\nA more explicit explanation of the relationship between NANOPUBLICATIONS and the overarching topic of PROVENANCE would be helpful.\n\"n-quads\" - please explain or provide a reference.\n\"At the time of writing, YummyData has collected data for about a year\" - what is the time of writing?\nCONCLUSION\n\"The NBDC RDF portal launched in 2015 complements the above resources...\" - again, this would be appropriate for a \"where are they now?\" section.\n\"will be addressed in the upcoming BioHackathons\" - they have been held annually since 2014, so it is odd to not mention what was addressed in subsequent events. And I hope those reports will be published in a more timely fashion!\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? No\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly", "responses": [] }, { "id": "67760", "date": "28 Aug 2020", "name": "João Moreira", "expertise": [ "Reviewer Expertise Ontology engineering", "semantic interoperability", "FAIR data principles", "e-Health semantics" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper presents an overview about experiences and produced work related to RDF from the 6th and 7th annual BioHackathons (2013 and 2014).\nThe paper is structured in two major sections about (1) RDF data management in life sciences; and (2) meta-data about the RDF datasets with emphasis on SPARQL endpoint resources. The paper provides an important contribution to the state-of-art in the context of RDF resources development and publication for life sciences, also presenting some discussion about good practices in these topics.\n\nThe paper is quite relevant and is well written, but there are some issues:\n\nMy main concern regards how current (actual) the content presented is, since the paper describes the editions of 2013 and 2014, and 7 editions occurred after these. I know that many advances/improvements happened in these last editions, including for example the use of ontologies like the Ontology for Biomedical Investigations (OBI), the Semanticscience Integrated Ontology (SIO), the HCLS ontology profile, the SPAR ontologies (particularly FaBiO), FHIR, among others, in different use cases and with more recent technologies (e.g., ShEx, SHACL, RDF*, GraphDB, JSON-LD, etc).\n\nIdeally, this survey should also incorporate these more recent BioHackathons (2015-2019) or at least provide a discussion about the main points where the most important advances/improvements happened. The current (recent) literature has many works produced in these last 7 editions.\n\nI missed a paragraph (or some sentences in the end of Introduction) explaining why the authors choose to structure the paper in these 2 major categories: (1) Improvement and utilization of RDF data in life sciences; and (2) Metadata about RDF data resources.\n\nThe reasoning behind the paper structure is not so clear. For example, it is difficult to understand why the sub-sections of \"Provenance of Data\" were organized in Nanopublication, Bio2RDF2SADI and Quality assessment. The 2 first categories refer to specific approaches, while the later refers to a generic data management activity. Similar issue happens on the other sections.\n\nThe discussion on best practices, outlined in the abstract as a contribution of this paper, is quite difficult to track. One can find some recommendations spread in the text, but, in my point of view, it would be better to have a section only for Discussion and Recommendations. So, I suggest to create this section before the conclusions, moving all good practices' discussion to it.\n\nFurthermore, it would be valuable to have in this new section a summary of the main open issues regarding the 2 major categories, highlighting which of the presented approaches faced these issues, giving directions to solve them.\n\nIn the Provenance of data section, I missed the reference to W3C PROV, which btw is used by the nanopublication approach and is a fundamental standard for reasoning over the data provenance.\n\nThe Conclusion section can be improved by making explicit what were the main lessons learned, the main contributions of the paper and a summary of the open issues identified.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-1677
https://f1000research.com/articles/8-1667/v1	20 Sep 19	{ "type": "Research Article", "title": "Gustatory dysfunction in relation to circumvallate papilla’s taste buds structure upon unilateral maxillary molar extraction in Wistar rats: an in vivo study", "authors": [ "Sana Mostafa", "Heba M. Hakam", "Amal El-motayam", "Heba M. Hakam", "Amal El-motayam" ], "abstract": "Background: The interaction between taste sensation and dentoalveolar innervation is still under research. teeth loss can alter taste thresholds in humans, but the underlying mechanisms are still obscure. This study investigated the effect of unilateral maxillary molars extraction on the structure of circumvallate papilla in rats. Methods: Thirty-two male Wister rats, aged 3-4 months were randomly distributed into four groups (one control and 3 experimental ) each including 8 animals. The rats were euthanized 3, 6 or 9 weeks following the procedure. The changes in trough length and the taste buds structure and number of both sides of CVP were investigated using routine histological examination followed by statistical analysis. Results: the trough toward the extraction side was obviously shorter with a noticeable decrease of taste buds’ number than the non-extraction side. Taste buds were reduced in size and most of them showed signs of degeneration which was more evident in group II followed by group III, less deformity detected in group IV in comparison to the preceding 2 experimental groups. the non-extraction side of all experimental groups showed normal trough length and generally normal histology of taste buds. Conclusions: Maxillary molars extraction has a degenerative effect on the structure of taste buds and gustatory epithelium which were more marked at the extraction side and showed improvement upon elongation of follow up period", "keywords": [ "Molars extraction", "taste sensation", "circumvallate papilla", "taste buds" ], "content": "Introduction\n\nTaste is an essential sensation that is mediated via taste buds (TBs) which are specialized neuroepithelial structures distributed predominantly over the dorsal surface of the tongue1–3, and undergo continuous renewal throughout adult life4,5.\n\nThe circumvallate papilla (CVP) is a major compartment of the gustatory system in the tongue that harbors more than 500 taste bud openings in its trough6,7. This gathering of taste organs in one specified place, make it easier to detect different changes under study.\n\nMoreover, the impact of nerve injury on fungiform taste buds appears to be less severe than on circumvallate taste buds, thus making it a good candidate to study neural variations8,9.\n\nThe gustatory system is unique in that taste buds innervation is essential to taste bud cell turnover10. In the rat tongue, the single central CVP is innervated by the glossopharyngeal nerve bilaterally7,11.\n\nTaste sensation is modulated by different factors including the integration of gustatory information with other sensory modalities such as olfaction, food texture, temperature and pain (hot spices) through the trigeminal nerve12. Other factors include nerve supply and mechanical stimuli that are affected by the integrity of masticatory apparatus13.\n\nTaste signals synapse to the nucleus of the solitary tract (NST) located in the medulla14. An interaction between dental and gustatory afferent neurons in the NST has been previously discussed9,15–17. In addition, teeth loss can alter taste thresholds in humans, but the underlying mechanisms are still obscure15,18,19.\n\nThis interaction has clinical implications as occlusion impairment by teeth extraction (dental deafferentation) can be considered a possible cause of taste alteration15.\n\nIt was reported that after unilateral transection of glossopharyngeal nerve in rat, the TBs loss in CVP was 12% rather than a 50% which indicate bilateral innervation of the remaining taste buds20. However, another study suggested that all the TBs in the rat CVP are bilaterally innervated except the upper two thirds of the outer trough walls which are innervated by ipsilateral glossopharyngeal nerve21.\n\nThe purpose of this study was to compare between both sides of the CVP regarding trough length and histopathological changes in the TBs structure and number secondary to unilateral maxillary molars extraction and to identify if unilateral extraction will cause damage to glossopharyngeal nerve on the same side of CVP only or may extend to the other side.\n\nRats were the animal of choice for this experiment because of their similarity to human anatomy regarding turnover time of taste buds, and innervation pattern of the tongue12,13.\n\n\nMethods\n\nA total of 32 male Wistar rats weighing 150–200 g, aged 3–4 months were used in this study. The animals were obtained from and housed in the Animal House, Faculty of Medicine, Cairo University. The animals were maintained in an air conditioned animal house under controlled room temperature 25±2°C with 12/12 h light/dark cycle and were allowed unlimited access to powdered soft food and water. Rats were housed in (standard polycarbonate cages (Pretty industries, Model:CR5). Control rats were placed as 4 animals per cage, while the experimental animals were placed one per cage until wounds had fully recovered in order to control infection.\n\nAll procedures were refined to reduce the negative impact on animals including administration of anesthesia prior to the procedure and gentle handling of the animal during extraction to prevent any tissue laceration or damage. After the experiment the animals received antibiotics to prevent infection, and analgesics to manage any pain or discomfort. The cages were cleaned frequently, they were supplied with wood chips substrate, colored tubes and ropes to ensure enriched environment for the animal.\n\nThe welfare of the animals was assessed prior to, during and after the experimental period by the attending veterinarian who intervened in case of pain or distress signs according to the ethical protocols for animal treatment that were supervised by the animal facilities, Faculty of Medicine, Cairo University. The study was approved by the Institutional Animal Care and Use Committee, Cairo University (Approval no. CUIIIS1016).\n\nSample size calculation based on previous research by Hsu et al., 2014; that reported the difference in the trough depth between the control and experimental (28 days) groups was 28 ±13 µm. Using power 90% and 5% significance level, 6 rats in each group was calculated to be sufficient to be able to reject the null hypothesis that the population means of the experimental and control groups are equal. That number was increased to 8 rats in each group to compensate for possible loss of animals during breeding. Sample size calculation were performed using PS: Power and Sample Size Calculation software Version 3.1.2 (Vanderbilt University, Nashville, Tennessee, USA).\n\nRats were randomly distributed by Random Sequence Generator program (randomizer.org) into four groups (one control and 3 experimental) each including 8 animals. Implementation of the allocation was done as follow: numbers from 1 to 32 were written on folded papers that were placed in opaque sealed envelopes, matching of the rats with the numbers were done blindly through the technician in charge at the animal house, each rat was attached to its number till the end, then the numbers were opened and the rats were allocated in their groups according to the program recommendations.\n\nThe control (group I) were anesthetized without tooth extraction (sham operation) and euthanized after 3 weeks. In the experimental groups, all maxillary molars of the right side were extracted (early in the morning) under general anesthesia by an intraperitoneal injection of ketamine/xylazine (ketamine 40–100 mg/kg IP/ xylazine 5–13 mg/kg IP, Trittau, Germany) for 60–80 minutes22. Extraction were performed using Halsted Curved Mosquito Hemostatic Forceps, Germany 15921-G with the tongue retracted by a tweezer. Postoperatively, antibiotics were administrated (cefotaxime10mg/Kg IV, Eipico, Egypt) for 3 days23. Analgesics were administrated (Ketoprofen (Anafen®, Merial) 5mg/kg I.M)24 once daily for up to 3 days.\n\nAnimals were euthanized with an intra-venous overdose of sodium thiopental (80mg/kg)25 (Anapental 1,P2-55-011, Sigma Tec Pharma –Egypt) (as a humane method that cause rapid euthanasia with minimal discomfort to the animals) after extraction at 3 weeks (group II), 6 weeks (group III) and 9 weeks (group IV) (Table 1).\n\nThe first euthanization date for the current study (3 weeks) were assigned according to Huang & Lu, 1996 study as they found that taste buds of the CVP were diminished 3 weeks after neurectomy26. The other intervals (6 and 9 weeks) were set in order to investigate the probability of recovery from any potential damage. The CVP was dissected out according to the animal’s tongue anatomy. The CVP is situated most posteriorly at the center of the base of the tongue, the site of the CVP in the dissected rat tongue was identified, then the tissue specimen was isolated by a horizontal cut anterior to the papilla and two oblique cuts posteriorly. Tissues were fixed in 10% calcium formol (No:8012-95-1, El-Gomhouria co., Egypt).\n\nSerial 5-microns thick sections were prepared using LEICA RM2255 microtome and picked up on glass slides (No: 7101, Yancheng Jingwei Chemicals Co., Ltd-china) for histological haematoxylin and eosin staining (No:L16220, L13960, El-Gomhouria co., Egypt).\n\nThe sections were deparaffinized and rehydrated by immersing them successively for ~5 min with agitation in xylene, 100% ethanol, and 70% ethanol. Then the slides were rinsed in running tap water at room temperature for at least 2 min. The sections were stained in hematoxylin solution for 3 min then placed under running tap water for 1 min. The sections were stained in 1% eosin Y solution for 2 min. The sections were dehydrated with two consecutive immersions in 95% alcohol, then two immersions in 100% alcohol for 30 sec each. The alcohol was then extracted with two consecutive immersions in xylene27. Blinding was applied for the primary animal care giver and the assessor of the histological and statistical results.\n\nThe specimens were examined using Leica DM300 light microscopic (Leica Microsystems, Inc., Switzerland).\n\nTBs’ number for each specimen were counted using an objective lens of magnification (x40) for all groups by light microscope. TBs were marked by green crosses and counted by Image J Version:1.52p computer system (Figure 1).\n\nThe length of both trough sides for each specimen was measured using an objective lens of magnification (x10) for all groups using light microscope. The length was measured using computer system (Figure 2).\n\nThe trough opening was determined by a solid line drawn between the 2 troughs, the trough depth was marked by two lines, from top to bottom.\n\nComparisons between groups were performed using one-way analysis of variance (ANOVA). This was followed by Tukey’s post hoc test for pairwise comparisons when ANOVA revealed a significant difference. Independent t test was used to compare both trough sides using IBM SPSS 18.0 version 21 for windows (SPSS Inc., Chicago, IL, USA). The significance level was set at p ≤ 0.05.\n\n\nResults\n\nAnalysis of the control group samples revealed normal CVP surrounded by even trough lengths with normal TB histology and number (Figure 3). In groups II and III, the papillae showed slight deformity of the general outline as the trough toward the extraction side was obviously shorter with noticeable decrease of TBs’ number than the non-extraction side. TBs were reduced in size and most of them showed signs of degeneration which was most evident in group II, followed by group III (Figure 4, Figure 5). A similar situation was observed in group IV with less deformity of the general outline, as the right side revealed an increase in length with apparent increase in TB number in comparison to the preceding two experimental groups. TBs exhibited nearly normal shape, with some degenerated taste cells (Figure 6). TBs of the non-extraction side in all experimental groups showed generally normal histology except for slight enlargement and few darkly stained nuclei (Figure 7). Raw microscope images are available as underlying data28–31.\n\n(A) Photomicrograph of control group (І) showing normal circumvallate papilla surrounded by a deep narrow trough of nearly equal length on both sides and gustatory epithelium with normal taste buds (H&E, orig. Mag.X100). (B) Higher magnification showing mature barrel shaped taste buds with taste pores, different taste cells types with prominent nuclei including supporting cells, gustatory receptor cells and basal cells (H&E, orig. Mag.X1000). H&E - haematoxylin and eosin, Mag - magnification.\n\n(A) Photomicrograph of group (ІI) showing deformity of the papilla general outline with shorter troughs toward the extraction side with fewer taste buds, and normal trough length toward the non-extraction side containing increased taste buds number (H&E, orig. Mag.X100). (B) Higher magnification showing immature taste buds toward the extraction side arranged at different levels and decreased in size with signs of degeneration (H&E, orig. Mag.X1000). H&E - haematoxylin and eosin, Mag - magnification.\n\n(A) Photomicrograph of group (ІII) showing shorter trough length toward the extraction side with decreased number of taste buds, focal area of inflammatory cells infiltration beneath the shorter trough and normal taste buds number in the other trough (H&E, orig. Mag.X100). (B) Higher magnification showing extraction side with decreased taste bud number with darkly stained nuclei (H&E, orig. Mag.X1000). H&E - haematoxylin and eosin, Mag - magnification.\n\n(A) Photomicrograph of group (ІV) showing trough toward the extraction side was nearly the same length as the other side, with increased number of taste buds in comparison to previous experimental groups (H&E, orig. Mag.X100). (B): Higher magnification of extraction side showing small size-buds with a few darkly stained nuclei (H&E, orig. Mag.X1000). H&E - haematoxylin and eosin, Mag - magnification.\n\nTaste buds of the non-extraction side in all experimental groups (A) group (II), (B) group (III) and (C) group (IV) showing generally normal histology except for slight enlargement (in groups II & III) and a few darkly stained nuclei (H&E, orig. Mag.X1000). H&E - haematoxylin and eosin, Mag - magnification.\n\nTB number and trough length variation between the two sides (extraction & non-extraction) were recorded among all experimental groups. It was noticed that non-extraction side recorded a higher mean number and length within each group, while the highest mean value was recorded in group II. This difference was not statistically significant in the controls, but was significant in groups II, III & IV (p ≤ 0.05). (Table 2 and Table 3, Figure 8 and Figure 9, underlying data32).\n\nSignificance level p<0.05, significant, NS=non significant\n\nSignificance level p<0.05, significant\n\n(blue columns for the extraction side and green columns for non-extraction side).\n\n(blue columns for the extraction side and green columns for non-extraction side).\n\n\nDiscussion\n\nTaste is a highly dynamic and complex process. The interaction between gustation and somatosensation of the tongue is still under investigation13. The consequences of dental deafferentation as one of the possible causes of taste disorders, is unclear and therefore an area for further investigation.\n\nThe current results are in accordance with Hsu et al., 2014 who observed decreased trough lengths with fewer TB numbers of the CVP 28 days following bilateral maxillary molars extraction in rats13. They suggested that this might be due to peripheral or central neuroplastic changes induced by dental deafferentation. Similarly, Braud et al., 2012 uncovered the interaction between some of dental afferents from the trigeminal nerve & gustatory neurons in the NST16.\n\nThe decreased TBs number toward the extraction side in the current study was in agreement with Huang & Lu, 1996 who studied unilateral glossopharyngeal nerve neurectomy, and concluded that the taste buds on the same side of the CVP decreased significantly in number during the first two weeks, and disappeared completely by week three26. Moreover, Ichimori et al., 2009 reported that denervation of gustatory epithelium causes apoptosis in all types of taste buds, resulting in rapid degeneration33.\n\nOn the other hand, Guth 1963 found that following unilateral glossopharyngeal nerve transection in the CVP of rats, the mean loss of taste buds in the outer trench walls was greater on the operated than the unoperated sides. However, they found that this difference was not statistically significant20. Guagliardo & Hill, 2007 demonstrated that TBs of fungiform papilla on the uncut side increased in size34. This might explain enlarged TBs on the non-extraction side.\n\nThe highest mean values were recorded in group II towards the non-extraction side, which could be explained as a compensatory mechanism for the damage that has occurred at the other side.\n\nImproved results observed in group IV regarding trough length and TBs’ number are probably due to the recovery of the gustatory epithelium with time. This is in agreement with Lim & Green, 2008 who postulated that nerve damage in certain region would augment remaining taste signals in the neighboring area35.\n\nStudy limitations may include that the process of counting TBs, which needs to be more precise, as the same taste bud may appear in different sections, like counting the total number of taste buds per papilla by scoring only TBs of maximum dimensions. A more accurate investigation methods could be used in further studies rather than mere TBs counting, like using reverse transcription polymerase chain reaction (RT-PCR) for detection of the expression of certain genes.\n\nThe current study has clinical implications, as it identified a correlation between normal occlusion and taste sensation, and it could be considered as an aid in explaining the etiology of taste loss. According to the current study results, it is important to screen subjects for dental status when planning taste studies, especially in sensory analyses. Subjects with recent multiple extractions should not be included in such studies as taste sensation is likely to have been affected. These findings also could help to distinguish the effect of normal aging from disease-related changes in taste perception.\n\nThe current study provides a step forward in predicting the effect of molar extraction and the impact of unilateral glossopharyngeal damage on taste sensation in humans. However, further investigations and observational studies are needed to prove that connection.\n\n\nConclusions\n\nMaxillary molars extraction has a degenerative effect on the structure of taste buds and the gustatory epithelium which were more marked at the extraction side, and showed improvement upon elongation of follow up period.\n\n\nData availability\n\nHarvard Dataverse: Dataset1. Control group (group I). https://doi.org/10.7910/DVN/UZ9OL028\n\nThis project contains the following underlying data:\n\nc1.tif (Microscope image, 10x magnification of control sample 1)\n\nc1a.tif (Microscope image, 40x magnification of control sample 1)\n\nc1b.tif (Microscope image, 40x magnification of control sample 1)\n\nc2.tif (Microscope image, 10x magnification of control sample 2)\n\nc2a.tif (Microscope image, 40x magnification of control sample 2)\n\nc2b.tif (Microscope image, 40x magnification of control sample 2)\n\nc3.tif (Microscope image, 10x magnification of control sample 3)\n\nc3a.tif (Microscope image, 40x magnification of control sample 3)\n\nc3b.tif (Microscope image, 40x magnification of control sample 3)\n\nc4.tif (Microscope image, 10x magnification of control sample 4)\n\nc4a.tif (Microscope image, 40x magnification of control sample 4)\n\nc4b.tif (Microscope image, 40x magnification of control sample 4)\n\nc5.tif (Microscope image, 10x magnification of control sample 5)\n\nc5a.tif (Microscope image, 40x magnification of control sample 5)\n\nc5b.tif (Microscope image, 40x magnification of control sample 5)\n\nc6.tif (Microscope image, 10x magnification of control sample 6)\n\nc6a.tif (Microscope image, 40x magnification of control sample 6)\n\nc6b.tif (Microscope image, 40x magnification of control sample 6)\n\nc7.tif (Microscope image, 10x magnification of control sample 7)\n\nc7a.tif (Microscope image, 40x magnification of control sample 7)\n\nc7b.tif (Microscope image, 40x magnification of control sample 7)\n\nc8.tif (Microscope image, 10x magnification of control sample 8)\n\nc8a.tif (Microscope image, 40x magnification of control sample 8)\n\nc8b.tif (Microscope image, 40x magnification of control sample 8)\n\ncontrol x1000a.tif (Microscope image, 100x magnification of control sample)\n\ncontrol x1000b.tif (Microscope image, 100x magnification of control sample)\n\nHarvard Dataverse: Dataset2. Group II samples. https://doi.org/10.7910/DVN/FOFV8930\n\nThis project contains the following underlying data:\n\ng2 (100x) ex..tif (Microscope image, 100x magnification of group II sample of extraction side)\n\ng2 (100x) ex.a.tif (Microscope image, 100x magnification of group II sample of extraction side)\n\ng2 (100x) ex.b.tif (Microscope image, 100x magnification of group II sample of extraction side)\n\ng2 (100x) non ex.a.tif (Microscope image, 100x magnification of group II sample of extraction side)\n\ng2 (100x) non ex.b.tif (Microscope image, 100x magnification of group II sample of extraction side)\n\ng2-1(10x).tif (Microscope image, 10x magnification of group II sample 1)\n\ng2-1(40x) ex..tif (Microscope image, 40x magnification of group II sample 1 of extraction side)\n\ng2-1(40x) non ex..tif (Microscope image, 40x magnification of group II sample 1 of non-extraction side)\n\ng2-2(10x).tif (Microscope image, 10x magnification of group II sample 2)\n\ng2-2(40x) ex..tif (Microscope image, 40x magnification of group II sample 2 of extraction side)\n\ng2-2(40x) non ex..tif (Microscope image, 40x magnification of group II sample 2 of non-extraction side)\n\ng2-3(10x).tif (Microscope image, 10x magnification of group II sample 3)\n\ng2-3(40x) ex..tif (Microscope image, 40x magnification of group II sample 3 of extraction side)\n\ng2-3(40x) non ex..tif (Microscope image, 40x magnification of group II sample 3 of non-extraction side)\n\ng2-4(10x).tif (Microscope image, 10x magnification of group II sample 4)\n\ng2-4(40x) ex..tif (Microscope image, 40x magnification of group II sample 4 of extraction side)\n\ng2-4(40x) ex..tif (Microscope image, 40x magnification of group II sample 4 of extraction side)\n\ng2-5(10x).tif (Microscope image, 10x magnification of group II sample 5)\n\ng2-5(40x) ex..tif (Microscope image, 40x magnification of group II sample 5 of extraction side)\n\ng2-5(40x) non ex..tif (Microscope image, 40x magnification of group II sample 5 of non-extraction side)\n\ng2-6(10x).tif (Microscope image, 10x magnification of group II sample 6)\n\ng2-6(40x) ex..tif (Microscope image, 40x magnification of group II sample 6 of extraction side)\n\ng2-6(40x) non ex..tif (Microscope image, 40x magnification of group II sample 6 of non-extraction side)\n\ng2-7(10x).tif (Microscope image, 10x magnification of group II sample 7)\n\ng2-7(40x) ex..tif (Microscope image, 40x magnification of group II sample 7 of extraction side)\n\ng2-7(40x) non ex..tif (Microscope image, 40x magnification of group II sample 7 of non-extraction side)\n\ng2-8(10x).tif (Microscope image, 10x magnification of group II sample 8)\n\ng2-8(40x) ex..tif (Microscope image, 40x magnification of group II sample 8 of extraction side)\n\ng2-8(40x) non ex..tif (Microscope image, 40x magnification of group II sample 8 of non-extraction side)\n\nHarvard Dataverse: Dataset3. Group III samples. https://doi.org/10.7910/DVN/KYNFDS30\n\nThis project contains the following underlying data:\n\ng3 (100x) ex.1.tif (Microscope image, 100x magnification of group III sample of extraction side)\n\ng3 (100x) ex.2.tif (Microscope image, 100x magnification of group III sample of extraction side)\n\ng3 (100x) ex.3.tif (Microscope image, 100x magnification of group III sample of extraction side)\n\ng3 (100x) non ex.1.tif (Microscope image, 100x magnification of group III sample of non-extraction side)\n\ng3 (100x) nonex.2.tif (Microscope image, 100x magnification of group III sample of non-extraction side)\n\ng3-1 (10x).tif (Microscope image, 10x magnification of group III sample 1)\n\ng3-1(40x) ex..tif (Microscope image, 40x magnification of group III sample 1 of extraction side)\n\ng3-1(40x) non ex..tif (Microscope image, 40x magnification of group III sample 1 of non-extraction side)\n\ng3-2 (10x).tif (Microscope image, 10x magnification of group III sample 2)\n\ng3-2(40x) ex..tif (Microscope image, 40x magnification of group III sample 2 of extraction side)\n\ng3-2(40x) non ex..tif (Microscope image, 40x magnification of group III sample 2 of non-extraction side)\n\ng3-3 (10x).tif (Microscope image, 10x magnification of group III sample 3)\n\ng3-3(40x) ex..tif (Microscope image, 40x magnification of group III sample 3 of extraction side)\n\ng3-3(40x) non ex..tif (Microscope image, 40x magnification of group III sample 3 of non- extraction side)\n\ng3-4 (10x).tif (Microscope image, 10x magnification of group III sample 4)\n\ng3-4(40x) ex..tif (Microscope image, 40x magnification of group III sample 4 of extraction side)\n\ng3-4(40x) non ex..tif (Microscope image, 40x magnification of group III sample 4 of non-extraction side)\n\ng3-5 (10x).tif (Microscope image, 10x magnification of group III sample 5)\n\ng3-5(40x) ex..tif (Microscope image, 40x magnification of group III sample 5 of extraction side)\n\ng3-5(40x) non ex..tif (Microscope image, 40x magnification of group III sample 5 of non-extraction side)\n\ng3-6 (10x).tif (Microscope image, 10x magnification of group III sample 6)\n\ng3-6(40x) ex..tif (Microscope image, 40x magnification of group III sample 6 of extraction side)\n\ng3-6(40x) non ex..tif (Microscope image, 40x magnification of group III sample 6 of non-extraction side)\n\ng3-7 (10x).tif (Microscope image, 10x magnification of group III sample 7)\n\ng3-7(40x) ex..tif (Microscope image, 40x magnification of group III sample 7 of extraction side)\n\ng3-7(40x) non ex..tif (Microscope image, 40x magnification of group III sample 7 of non-extraction side)\n\ng3-8 (10x).tif (Microscope image, 10x magnification of group III sample 8)\n\ng3-8(40x) ex..tif (Microscope image, 40x magnification of group III sample 8 of extraction side)\n\ng3-8(40x) non ex..tif (Microscope image, 40x magnification of group III sample 8 of non-extraction side)\n\nHarvard Dataverse: Dataset4. Group IV samples. https://doi.org/10.7910/DVN/OLZF0L31\n\nThis project contains the following underlying data:\n\ng4 (100x) ex.1.tif (Microscope image, 100x magnification of group IV sample of extraction side)\n\ng4 (100x) ex.2.tif (Microscope image, 100x magnification of group IV sample of extraction side)\n\ng4 (100x) non ex.1.tif (Microscope image, 100x magnification of group IV sample of non-extraction side)\n\ng4 (100x) non ex.2.tif (Microscope image, 100x magnification of group IV sample of non-extraction side)\n\ng4-1 (10x).tif (Microscope image, 10x magnification of group IV sample 1)\n\ng4-1(40x) ex..tif (Microscope image, 40x magnification of group IV sample 1 of extraction side)\n\ng4-1(40x) non ex..tif (Microscope image, 40x magnification of group I V sample 1 of non-extraction side)\n\ng4-2 (10x).tif (Microscope image, 10x magnification of group IV sample 2)\n\ng4-2(40x) ex..tif (Microscope image, 40x magnification of group IV sample 2 of extraction side)\n\ng4-2(40x) non ex..tif (Microscope image, 40x magnification of group IV sample 2 of non-extraction side)\n\ng4-3 (10x).tif (Microscope image, 10x magnification of group IV sample 3)\n\ng4-3(40x) ex..tif (Microscope image, 40x magnification of group IV sample 3 of extraction side)\n\ng4-3(40x) non ex..tif (Microscope image, 40x magnification of group IV sample 3 of non- extraction side)\n\ng4-4 (10x).tif (Microscope image, 10x magnification of group IV sample 4)\n\ng4-4(40x) ex..tif (Microscope image, 40x magnification of group IV sample 4 of extraction side)\n\ng4-4(40x) non ex..tif (Microscope image, 40x magnification of group IV sample 4 of non-extraction side)\n\ng4-5 (10x).tif (Microscope image, 10x magnification of group IV sample 5)\n\ng4-5(40x) ex..tif (Microscope image, 40x magnification of group IV sample 5 of extraction side)\n\ng4-5(40x) non ex..tif (Microscope image, 40x magnification of group IV sample 5 of non-extraction side)\n\ng4-6 (10x).tif (Microscope image, 10x magnification of group IV sample 6)\n\ng4-6(40x) ex..tif (Microscope image, 40x magnification of group IV sample 6 of extraction side)\n\ng4-6(40x) non ex..tif (Microscope image, 40x magnification of group IV sample 6 of non-extraction side)\n\ng4-7 (10x).tif (Microscope image, 10x magnification of group IV sample 7)\n\ng4-7(40x) ex..tif (Microscope image, 40x magnification of group IV sample 7 of extraction side)\n\ng4-7(40x) non ex..tif (Microscope image, 40x magnification of group IV sample 7 of non-extraction side)\n\ng4-8 (10x).tif (Microscope image, 10x magnification of group IV sample 8)\n\ng4-8(40x) ex..tif (Microscope image, 40x magnification of group IV sample 8 of extraction side)\n\ng4-8(40x) non ex..tif (Microscope image, 40x magnification of group IV sample 8 of non-extraction side)\n\nHarvard Dataverse: main manuscript data. https://doi.org/10.7910/DVN/9IA7CS32\n\nThis project contains the following underlying data:\n\ntrough length.tab (Excel spreadsheet of measured trough lengths)\n\ntaste buds number.tab (Excel spreadsheet of recorded taste bud numbers)\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "References\n\nChaudhari N, Roper SD: The cell biology of taste. J Cell Biol. 2010; 190(3): 285–296. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFernstrom JD, Munger SD, Sclafani A, et al.: Mechanisms for sweetness. J Nutr. 2012; 142(6): 1134S–1141S. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIwasaki S, Yoshizawa H, Kawahara I: Study by scanning electron microscopy of the morphogenesis of three types of lingual papilla in the rat. Anat Rec. 1997; 247(4): 528–541. PubMed Abstract \| Publisher Full Text\n\nBarlow LA: Progress and renewal in gustation: new insights into taste bud development. Development. 2015; 142(21): 3620–3629. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFeng P, Huang L, Wang H: Taste bud homeostasis in health, disease, and aging. Chem Senses. 2014; 39(1): 3–16. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBradley RM, Mistretta CM, Nagai T: Demonstration of sensory innervation of rat tongue with anterogradely transported horseradish peroxidase. Brain Res. 1986; 367(1–2): 364–367. PubMed Abstract \| Publisher Full Text\n\nOakley B, Witt M: Building sensory receptors on the tongue. J Neurocytol. 2004; 33(6): 631–646. PubMed Abstract \| Publisher Full Text\n\nGuagliardo NA, Hill DL: Fungiform taste bud degeneration in C57BL/6J mice following chorda-lingual nerve transection. J Comp Neurol. 2007; 504(2): 206–216. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMiura H, Scott JK, Harada S, et al.: Sonic hedgehog-expressing basal cells are general post-mitotic precursors of functional taste receptor cells. Dev Dyn. 2014; 243(10): 1286–1297. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMeng L, Ohman-Gault L, Ma L, et al.: Taste Bud-Derived BDNF Is Required to Maintain Normal Amounts of Innervation to Adult Taste Buds. eNeuro. 2015; 2(6): 1–20, pii: ENEURO.0097-15.2015. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMistretta CM, Goosens KA, Farinas I, et al.: Alterations in size, number, and morphology of gustatory papillae and taste buds in BDNF null mutant mice demonstrate neural dependence of developing taste organs. J Comp Neurol. 1999; 409(1): 13–24. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYarmolinsky DA, Zuker CS, Ryba NJ: Common sense about taste: from mammals to insects. Cell. 2009; 139(2): 234–244. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHsu JC, Watari I, Ono R, et al.: Degeneration of fungiform and circumvallate papillae following molar extraction in rats. Acta Odontol Scand. 2014; 72(8): 880–886. PubMed Abstract \| Publisher Full Text\n\nPurves D, Augustine GJ, Fitzpatrick D, et al.: Neuroscience. 2004; 3. doi:978-0878937257.\n\nBoucher Y, Berteretche MV, Farhang F, et al.: Taste deficits related to dental deafferentation: an electrogustometric study in humans. Eur J Oral Sci. 2006; 114(6): 456–464. PubMed Abstract \| Publisher Full Text\n\nBraud A, Vandenbeuch A, Zerari-Mailly F, et al.: Dental afferents project onto gustatory neurons in the nucleus of the solitary tract. J Dent Res. 2012; 91(2): 215–220. PubMed Abstract \| Publisher Full Text\n\nAvivi-Arber L, Seltzer Z, Friedel M, et al.: Widespread Volumetric Brain Changes following Tooth Loss in Female Mice. Front Neuroanat. 2017; 10: 121. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMeng L, Jiang X, Ji R: Role of neurotrophin in the taste system following gustatory nerve injury. Metab Brain Dis. 2015; 30(3): 605–613. PubMed Abstract \| Publisher Full Text\n\nShafer DM, Frank ME, Gent JF, et al.: Gustatory function after third molar extraction. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 1999; 87(4): 419–428. PubMed Abstract \| Publisher Full Text\n\nGuthi L: Histological changes following partial denervation of the circumvallate papilla of the rat. Exp Neurol. 1963; 8(4): 336–349. Publisher Full Text\n\nState FA: Histological changes following unilateral re-innervation of the circumvallate papilla of rat. Acta Anat (Basel). 1977; 98(3): 343–352. PubMed Abstract \| Publisher Full Text\n\nIACUC Anesthesia Guideline at the University of Iowa. Reference Source\n\nCoombes JD: Metabolism of cefotaxime in animals and humans. Rev Infect Dis. 1982; 4 Suppl: S325–S332. PubMed Abstract \| Publisher Full Text\n\nUBC Animal Care Guidelines: SOP: UBC ACC TECH 18 Ketoprofen.\n\nClose B, Banister K, Baumans V, et al.: Recommendations for euthanasia of experimental animals: Part 2. DGXT of the European Commission. Lab Anim. 1997; 31(1): 1–32. PubMed Abstract \| Publisher Full Text\n\nHuang YJ, Lu KS: Unilateral innervation of guinea pig vallate taste buds as determined by glossopharyngeal neurectomy and HRP neural tracing. J Anat. 1996; 189(Pt 2): 315–324. PubMed Abstract \| Free Full Text\n\nFischer AH, Jacobson KA, Rose J, et al.: Hematoxylin and eosin staining of tissue and cell sections. CSH Protoc. 2008; 2008: pdb.prot4986. PubMed Abstract \| Publisher Full Text\n\nMoustafa Hussein, Sanha: \"dataset1\". Harvard Dataverse, V1. 2019. http://www.doi.org/10.7910/DVN/UZ9OL0\n\nMoustafa Hussein, Sanha: \"dataset2\". Harvard Dataverse, V1. 2019. http://www.doi.org/10.7910/DVN/FOFV89\n\nMoustafa Hussein, Sanha: \"dataset 3\". Harvard Dataverse, V1. 2019. http://www.doi.org/10.7910/DVN/KYNFDS\n\nMoustafa Hussein, Sanha: \"dataset 4\". Harvard Dataverse, V1. 2019. http://www.doi.org/10.7910/DVN/OLZF0L\n\nAlmogheer, Sana: \"main manuscript data\". Harvard Dataverse, V2. 2019. http://www.doi.org/10.7910/DVN/9IA7CS\n\nIchimori Y, Ueda K, Okada H, et al.: Histochemical changes and apoptosis in degenerating taste buds of the rat circumvallate papilla. Arch Histol Cytol. 2009; 72(2): 91–100. PubMed Abstract \| Publisher Full Text\n\nGuagliardo NA, Hill DL: Fungiform taste bud degeneration in C57BL/6J mice following chorda-lingual nerve transection. J Comp Neurol. 2007; 504(2): 206–216. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLim J, Green BG: Tactile interaction with taste localization: influence of gustatory quality and intensity. Chem Senses. 2008; 33(2): 137–143. PubMed Abstract \| Publisher Full Text" }	[ { "id": "54179", "date": "07 Oct 2019", "name": "Mahmoud M. Al-Ankily", "expertise": [ "Reviewer Expertise Oral biology", "oral histology", "immunohistochemistry", "oral anatomy and dentistry" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article by Sana Mostafa et al. examines Gustatory dysfunction in relation to circumvallate papilla’s taste buds structure upon unilateral maxillary molar extraction in Wistar rats. The study, although it is simple, adds knowledge to the literature.\nMinor comments would help to improve the impact of this paper:\nTitle: Gustatory dysfunction in relation to circumvallate papilla’s taste buds structure upon unilateral maxillary molar extraction in Wistar rats but in Methods you said (all maxillary molars of the right side were extracted) so you may change it to maxillary molars.\nIntroduction: An interaction between dental and gustatory afferent neurons in the NST has been previously discussed please add a hint about it\nMethods: In the experimental groups, all maxillary molars of the right side were extracted (early in the morning) why? Is the time of extraction make a significant difference.\nThe tongue retracted by a tweezer. Postoperatively, antibiotics were administrated (cefotaxime10mg/Kg IV, Eipico, Egypt) for 3 days23. Analgesics were administrated (Ketoprofen (Anafen®, Merial) 5mg/kg I.M)24 once daily for up to 3 days did you do that with the control group? If not how to be sure that the medications does not affect the experiment.\nResults: Figures from 2 to 6: please clarify which side is the extracted molars side and adjust the angulation of the CVP (figure 3 and 5) to make it easy to compare between the 2 sides of the trough.\nDiscussion Further investigations of the nerve supply under TBs of both sides using immunohistochemistry is recommended.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "57946", "date": "07 Jan 2020", "name": "Mahmoud M. Bakr", "expertise": [ "Reviewer Expertise Oral Biology" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAbstract: Some typos were evident, first line (teeth) should start with an upper case T as it is a new sentence. The same applies to (the) at the beginning of the results section.\nAlso in the results section the authors report findings from groups II, III and IV without any previous definition of what each group consist of. The abstract should be readable as a separate entity without having to refer to the full article.\nIn conclusions (which were more marked) should read (which was more marked) as it refers to the degenerative effect. Alternatively, the first sentence in the conclusions section should read \"Maxillary molars extraction has degenerative effects....\nIntroduction: This section is well written but needs to be structured and reformatted into paragraphs rather than isolated sentences.\nMethods: First paragraph, there is an extra bracket () before standard polycarbonate cages that needs to be deleted.\nResults: Describing deformity of taste buds is very subjective and a more objective way of defining the degree of deformity is needed.\nFigure 4A is of a very poor quality and there are some concerns that it does not belong to this experiment at all due to different light conditions and degree/type of stain. It seems like a texbook image rather than an experimental one. Please replace this image.\nUnfortunately, nothing was reported about the differences between the four groups using a one-way ANOVA. Even if there were no significant findings, it should still be reported.\nFor the independent t-test results, please report the actual P value and standard deviations rather than (P<0.05). This is important to inform the readers about the degree of variability within your samples and the level of statistical significance.\nDiscussion: It is recommended to discuss the difference that extraction of lower molars and wisdom teeth (in humans) would make to taste sensation as well.\nConclusions: It is misleading to report that extraction of maxillary molars had a degenerative effect on circumvallate papillae in general as no data from this study supports this statement. The data from this study only shows degeneration in comparison to the non-extraction side. Also the claim that the degeneration improves by time is only an observation and is not supported by statistical analysis (No data from the one way ANOVA) was presented.\nThe conclusions need to rewritten both in the abstract and the full article text as it is over presenting the limited findings obtained from this study.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-1667
https://f1000research.com/articles/8-975/v1	27 Jun 19	{ "type": "Opinion Article", "title": "Pathology and regulation for research in the UK: an overview", "authors": [ "Owen John Driskell", "Jessica L. Lee", "Karin A. Oien", "Andy Hall", "Clare Verrill", "CM-Path QA Panel", "Jessica L. Lee", "Karin A. Oien", "Andy Hall", "Clare Verrill" ], "abstract": "The input of pathologists is essential for the conduct of many forms of research, including clinical trials. As the custodians of patient samples, pathology departments have a duty to ensure compliance with the relevant regulations, standards and guidelines to ensure the ethical and effective use for their intended investigational analysis. This includes where patients are participating in a research study. The results of research studies have impacts beyond the research study itself as they may inform changes in policy and practice or support the licensing of medicines and devices. Compliance with regulations and standards provides public assurance that the rights, safety and wellbeing of research participants are protected, that the data have been collected and processed to ensure their integrity and that the research will achieve its purpose. The requirements of the regulatory environment should not be seen as a barrier to research and should not significantly impact on the work of the laboratory once established and integrated into practice. This paper highlights important regulations, policy, standards and available guidance documents that apply to research involving NHS pathology departments and academic laboratories that are contributing to research involving human subjects.", "keywords": [ "Research regulation", "Research governance", "MHRA", "GCP", "HTA", "Quality Management", "Clinical Trial", "CTIMP" ], "content": "Introduction\n\nEvidence from research is needed to justify changes in practice and is a driving force behind improvements in patient care. Pathology can be involved throughout the research pathway - from developing research ideas, designing, setting up and running research studies, to analysing samples, interpreting the data and publishing the results (Figure 1).\n\nThe pathology laboratory has different potential functions along the study pathway. Pathologists may be leading the Research Generation (green), e.g. coming up with the research question, applying for funding, writing up the research or providing pathology expertise to these activities as part of a multidisciplinary team. Pathologists may also be responsible for providing supporting Research Delivery activities (blue) for studies from outside researchers, e.g. assessing research protocols for local delivery and providing service activities such as reporting on or arranging the release of diagnostic tissue to research centres.\n\nResearch may be initiated by pathologists, but a large amount of research is led by researchers from outside pathology. This research often requires services delivered by pathology departments. For some pathology staff this service delivery, for example the release of material from hospital archives, may be their only experience of research. Data generated from pathology departments supports research in a wide variety of ways. This could be informing study eligibility assessments, companion diagnostics, pharmacokinetics, surrogate or other endpoints or the provision of data, results or tissue for further assessment. Results may already be in the health record as part of routine care or they may require further analysis or the release of tissue to central laboratories for a standardised assessment.\n\nAs well as the increase in the overall workload of pathology departments, there have been increases in clinical trials requiring pathology input. As research needs to be carried out in accordance with the appropriate regulatory requirements and standards, pathology departments need to have risk-based and pragmatic procedures in place to manage these requests. Regulations change and so it is important that pathology departments maintain an understanding of the current regulatory environment around research.\n\n\nUK Policy Framework for Health and Social Care Research\n\nIn the NHS research should be carried out in accordance with the UK Policy Framework for Health and Social Care Research1, which replaced the Department of Health Research Governance Framework in October 2017. The Framework sets out the principles of best practice for the management and conduct of all health and social care research in the UK, to help organisations meet their legal requirements. It covers research in the NHS and other health and social care environments across all four nations. The Framework includes fifteen principles that apply to all research and a further four that apply only to interventional research. It provides a clear definition and outline of the roles and responsibilities for individuals and organisations involved in health and social care research including the sponsor, the funder, the research team, research sites, clinical research organisations and regulators. It states there should be clear designation of responsibility and accountability with clear lines of communication between all those involved in research.\n\nThe Framework is not a regulation - it is statutory guidance published in order to help organisations meet their legal obligations. It is a common set of principles that apply to any study. Though they aren’t identical, these will be recognisable to those already familiar with the principles of Good Clinical Practice. The framework is an important document for all those involved in NHS research as it provides a common language for research and enhances the possibilities for a unified and streamlined approach to research delivery.\n\n\nGood Clinical Practice, the UK Clinical Trials Regulations and the MHRA\n\nThe International Council for Harmonisation (ICH) of Technical Requirements for Pharmaceuticals for Human Use Good Clinical Practice (GCP) is an international standard for the design, conduct, performance, monitoring, auditing, recording, analysis, and reporting of clinical trials. These internationally recognised standards followed on from the Nuremberg Code and originate from the World Medical Association Declaration of Helsinki 1964 and were an attempt to overcome international inconsistencies in GCP2.\n\nThe Guideline for Good Clinical Practice E6 from the European Medicines Agency includes the 13 Principles of ICH GCP3. These guidelines refer to ‘clinical trials’ but may also be applied to other clinical investigations that may have an impact on the safety and well-being of human participants.\n\nICH GCP was incorporated into EU Directives 2001/20/EC and 2005/28/EC aimed at harmonising the regulation of Clinical Trials in the EU4,5. These were transposed into UK Law via The Medicines for Human Use (Clinical Trials) Regulations in 20046 revised in 20067 and have since been further amended. The UK has adopted the principles of GCP, but an additional principle relating to sponsor indemnity was added. This did not define ICH GCP as the legal standard unlike other European countries. As with the UK Policy Framework for Health and Social Care Research these regulations also include definitions and roles and responsibilities for research.\n\nThe Clinical Trial regulations define a clinical trial of an investigational medicinal product (CTIMP) as any investigation in human subjects, other than a non-interventional trial, intended (a) to discover or verify the clinical, pharmacological or other pharmacodynamic effects of one or more medicinal products, (b) to identify any adverse reactions to one or more such products, or (c) to study absorption, distribution, metabolism and excretion of one or more such products, with the object of ascertaining the safety or efficacy of those products. The Medicines and Healthcare products Regulatory Agency (MHRA) provide a tool for whether research is a CTIMP8 and can provide further assistance if required.\n\nThe regulations have had subsequent amendments in response to developments in healthcare, for example research interventions in pandemic disease and advanced therapies such as gene therapy and tissue engineering. Europe will be implementing a new Clinical Trial Regulation (536/2014)9 during 2020 and this may be implemented in the UK dependent on Brexit negotiations and future agreements. If the new Regulation does not come into force during the implementation period, the Government has confirmed that UK law will remain aligned with parts of the EU’s CTR legislation that are within the UK’s control, in order that researchers conducting clinical trials can plan with greater certainty10.\n\nThe Medicines for Human Use Regulations permit inspection of clinical trials and their conduct, including work undertaken in a laboratory, by the MHRA GCP Inspectorate. It is therefore prudent for a laboratory to know which trials they support are CTIMPs and so are subject to these regulations and potential inspection.\n\nThe primary reference for describing the requirements of GCP in the context of laboratories is the EMA Reflection Paper for Laboratories that Perform the Analysis or Evaluation of Clinical Trial Samples published in 201211. This was developed based on the MHRA Guidance on the Maintenance of Regulatory Compliance in Laboratories that Perform the Analysis or Evaluation of Clinical Trial Samples (published in 2009 now withdrawn). This document describes expectations for laboratories involved in the analysis of samples originating from a CTIMP.\n\nThe primary focus of inspection by the MHRA GCP Inspectorate is the analysis of samples that support primary and secondary endpoints and objectives of the trial. Routine sample analyses for safety testing within clinical chemistry and haematology laboratories are not the central focus of this inspection programme. The legislation does not differentiate between the purpose of different types of sample analysis; this is a pragmatic approach by the MHRA Inspectorate and potentially any aspects may be reviewed during an inspection. It is expected that laboratories consider and implement the guidance to support all forms of sample analysis linked to clinical trials whilst ensuring that this is appropriate and proportionate for the work being undertaken.\n\nFurther information in relation to the conduct of clinical trials can be found in the Good Clinical Practice Guide (otherwise known as the “Grey Guide”) produced by the MHRA12. This covers all aspects of GCP and is a recommended reference for laboratories involved in clinical research (including those not involved in primary and secondary endpoint analysis). The MHRA also continues to publish blogs on a variety of topics including those directly relevant to laboratory involvement in CTIMPs13.\n\n\nHuman Tissue Act 2004 and the Human Tissue Authority\n\nThe Human Tissue Act (2004) relates to the removal, storage and use of ‘relevant material’ for Scheduled Purposes14. The Act makes consent the fundamental principle to the use of human tissue. Although there are caveats, relevant material generally refers to samples, other than gametes, which include intact human cells. This includes most types of samples held in histology departments. Samples rendered acellular are not considered to be relevant material. There are eleven Scheduled Purposes for which consent is required, including ‘Research in connection with disorders, or the functioning, of the human body’. Samples collected as part of routine clinical care are not covered by the Human Tissue Act though they could subsequently be used for a Scheduled Purpose. The Act only applies in England, Wales and Northern Ireland as Scotland has separate legislation which only applies to samples collected from the deceased15.\n\nThe Human Tissue Authority (HTA) was created by the Human Tissue Act to regulate the removal, storage and use of human tissue. They grant licences to establishments for a range of specific activities involving bodies and relevant material, including undertaking post-mortem examinations and storage of relevant material for Scheduled Purposes. HTA-licensed establishments include hospitals (both NHS and private), tissue banks, NHS blood and transfusion premises, private pathology services, clinical research facilities, commercial life sciences laboratories and universities.\n\nInstitutions handling human samples must ensure that they have any necessary licences for the activities carried out. The HTA does not license the ‘use’ of tissue for research or approve individual research projects or clinical trials. Neither do they have a role in the ethical approval of research. A research sector licence covers the storage of the relevant material. Storage of relevant material for research may be exempt from the need for a HTA licence if the sample is being stored according to the protocol of a research study approved by an NHS Research Ethics Committee. The HTA standards and guidance for research are contained within Code A: Guiding Principles and the fundamental principle of consent16 and Code E: Research17. There is a useful decision chart for whether you need a research licence on the HTA website (Code E Annex C).\n\n\nOther regulation\n\nThe regulations and guidelines that relate to research do not replace the existing regulations and guidelines that govern laboratory practice. All the existing professional standards and regulations still apply and those that relate specifically to research must be applied in addition to these. Examples include ISO-1518918, the Data Protection Act19 implementing the General Data Protection Directive (GDPR) (as well as being custodian of patient samples Pathology is also custodian of their data) and the Transport of Dangerous Goods regulations20. Researchers must be aware of, and adhere to, local NHS Trust policies and relevant professional standards such as those set by the General Medical Council (GMC), Royal College of Pathologists, the Institute of Biomedical Science (IBMS) and the Association of Clinical Scientists (ACS). These professional standards include the need to be appropriately qualified for their job roles and the need to maintain continued professional development mirroring requirements in research regulation. Further, international trials may be subject to the scrutiny of other agencies such as the US Food and Drug Administration (FDA) so it is important to consider differing regulatory requirements depending on the intended use of the data.\n\n\nAccreditation and research\n\nNHS Pathology laboratories and those laboratories who are providing clinical services are used to what was Clinical Pathology Accreditation and is now Accreditation to ISO 15189 (Medical Laboratories - requirements for quality and competence in medical laboratories)18 through the United Kingdom Accreditation Service (UKAS). These standards are about Quality Management supporting a culture and systems that can also be used to support the needs of research regulations and standards.\n\nThere are no laboratory accreditation schemes for the analysis of clinical trial samples run by the MHRA. For CTIMPs, laboratories are required to be compliant with all relevant legislation which includes working to the principles of GCP. Where analysis is undertaken for the purposes of patient safety during a CTIMP and is part of a routine repertoire of practice, UKAS accreditation is taken into account. The regulations permit this work to be inspected but often there are other activities which are prioritised for inspection due to the risk-based approach taken. Study sponsors frequently ask for the UKAS accreditation status of a clinical laboratory as part of their study set up process. However, ISO 15189 itself makes no specific mention of research. While the standards are complementary, laboratory practice also needs to quality manage areas of research practice not covered by ISO 15189. Examples include provision for the management of result reporting in CTIMP studies where blinding may be compromised, data integrity considerations21 and the strict processing of trial samples according to the trial protocol.\n\nLaboratories can be accredited to Good Clinical Laboratory Practice (GCLP) by third party providers. GCLP is a term for various guidance documents with their origins in a set of standards produced by the World Health Organisation (WHO) for the purposes of supporting safety and efficacy in international studies in developing countries22 and subsequently reproduced by a number of non-government organisations23. Good Laboratory Practice (GLP) applies to all non-clinical safety studies which are designed to determine the effects of a chemical on human health, animal health and the environment. GLP specifically excludes clinical experiments. Both GLP and GCP are statutory requirements. In contrast, adherence to this GCLP is not a statutory requirement and is not assessed during regulatory GCP laboratory inspections performed by organisations like the MHRA.\n\n\nConclusion\n\nThere has been an increased emphasis on the importance of research to the NHS. Pathology’s part in healthcare and research is set to become even more important with diagnostics being central to the rise in genomic, personalised and stratified medicine. It is important that laboratories can demonstrate they have oversight of the research activity undertaken in their department. Quality management, with clear and recognisable processes, record keeping (and therefore traceability) and good communication all support compliance with regulation and guidance. The regulation and guidance around research continues to evolve, reflecting developments in healthcare. It is important that pathologists remain informed about the regulatory requirements, standards and guidance and maintain their professionalism towards the research activities they are involved with. This paper is an overview to highlight some of the regulations and policies around research rather than a comprehensive review. Digital and molecular pathology are areas of research shaping future clinical practice. Through such continued leadership in research, pathology is and will remain a central scientific discipline in clinical practice.\n\n\nData availability\n\nNo data are associated with this article.", "appendix": "Grant information\n\nThe CM-Path initiative and QA Panel is supported by the National Cancer Research Institute (NCRI).\n\nOJD is funded by the National Institute for Health Research (NIHR) Clinical Research Network West Midlands and is a member of the CM-Path Clinical Trials Working Group supported by the NCRI. CV is funded by the Oxford NIHR Biomedical Research Centre.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe would like to thank Jason Wakelin-Smith from the GCP & GLP Inspectorate, at the Medicines and Healthcare products Regulatory Agency and Christopher Birkett from the Human Tissue Authority for their helpful comments and suggestions in the preparation of this paper.\n\nA list of CM-Path QA Panel members is available here.\n\n\nReferences\n\nUK Policy Framework for Health and Social Care Research. (Accessed: April 2019). Reference Source\n\nVijayananthan A, Nawawi O: The importance of Good Clinical Practice guidelines and its role in clinical trials. Biomed Imaging Interv J. 2008; 4(1): e5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGuideline for good clinical practice E6 (R2). (EMA/CHMP/ICH/135/1995). (Accessed: April 2019). Reference Source\n\nDIRECTIVE 2001/20/EC OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL. (Accessed: April 2019). Reference Source\n\nDIRECTIVE 2005/28/EC OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL. (Accessed: April 2019). Reference Source\n\nThe Medicines for Human Use (Clinical Trials) Regulations 2004. (Accessed: April 2019). Reference Source\n\nThe Medicines for Human Use (Clinical Trials) Amendment Regulations 2006. (Accessed: April 2019). Reference Source\n\nMHRA CTIMP decision tool. (Accessed: April 2019). Reference Source\n\nREGULATION (EU) No 536/2014 OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL. (Accessed: April 2018). Reference Source\n\nTechnical information on what the implementation period means for the life science sector. (Accessed: April 2019). Reference Source\n\nReflection paper for laboratories that perform the analysis or evaluation of clinical trial samples (EMA/INS/GCP/532137/2010). European Medicines Agency. (Accessed: April 2019). Reference Source\n\nGood Clinical Practice Guide: MHRA, 2012. (publisher ‘The Stationary Office’), IBSN: 9780117081079. Reference Source\n\nMaking GCP Training Relevant and Applicable: It’s Not Just for Clinical Staff. Jason Wakelin-Smith. 2018; (Accessed: April 2019). Reference Source\n\nHuman Tissue Act 2004. (Accessed: April 2019). Reference Source\n\nHuman Tissue (Scotland) Act 2006. (Accessed: April 2019). Reference Source\n\nCode A: Guiding principles and the fundamental principle of consent (HTA). (Accessed: April 2019). Reference Source\n\nCode E: Research (HTA). (Accessed: April 2019). Reference Source\n\nISO 15189:2012(en) Medical laboratories - requirements for quality and competence. (Accessed: April 2019). Reference Source\n\nData Protection Act 2018. (Accessed: April 2019). Reference Source\n\nThe Carriage of Dangerous Goods and Use of Transportable Pressure Equipment Regulations 2009. (Accessed: April 2019). Reference Source\n\nMHRA ‘GXP’ Data Integrity Guidance and Definitions. (Accessed: April 2019). Reference Source\n\nGood Clinical Laboratory Practice. WHO, 2009. (Accessed: April 2019). Reference Source\n\nGood Clinical Laboratory Practices. RQA, 2012. (Accessed: April 2019). Reference Source" }	[ { "id": "50493", "date": "11 Jul 2019", "name": "Mark Sobel", "expertise": [ "Reviewer Expertise Biomedical research ethics", "biobanking", "molecular pathology", "neoplasia" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe Opinion Article by Owen Driskell and colleagues cites and briefly explains the relevant regulations in the UK concerning research conducted in pathology departments, with an emphasis on clinical trials. Editorially, and consistent with the principles of the UK policy framework for health and social care research (reference 1 - https://www.hra.nhs.uk/planning-and-improving-research/policies-standards-legislation/uk-policy-framework-health-social-care-research/) the authors state that the “requirements of the regulatory environment should not be seen as a barrier to research and should not significantly impact on the work of the laboratory once established and integrated into practice.”\nThe article is organized, generally well written and succinct, and provides links (in the references) to the appropriate statutes and regulations.\nThis reviewer recommends (in the section “Human Tissue Act 2004 and the Human Tissue Authority”) that the authors clarify whether the Human Tissue Act covers analytes stored in pathology departments such as DNA, RNA, and protein that have been isolated from human tissues. The Human Tissue Act does not consider acellular samples to be “relevant material;” however the Act clearly covers the use of DNA in research.\nEditing suggestions for consideration by the authors:\nAbstract (sentences 2-3): “As the custodians of patient samples, pathology departments have a duty to ensure compliance with the relevant regulations, standards and guidelines to ensure the ethical and effective use for their intended investigational analysis, including when patients are participating in a research study.”\n\nParagraph 2 (sentences 1-2): “In addition to research initiated by pathologists, a large amount of research is led by researchers from outside pathology requiring services delivered by pathology departments.”\n\nParagraph 2 (sentence 4): “Data generated from pathology departments support research in a wide variety of ways.” (“Data” is plural.)\n\nParagraph 13 (beginning “The primary focus of inspection by the MHRA…”) the second and third sentences could be misinterpreted. I suggest rewording as follows: “Although routine sample analyses for safety testing within clinical chemistry and haemotology laboratories are not the central focus of this inspection programme, the legislation does not differentiate between the purpose of different types of sample analysis; this is a pragmatic approach by the MHRA Inspectorate and potentially any aspects may be reviewed during an inspection.”\n\nParagraph 17 (sentence 2): All the existing professional standards and regulations still apply and those that relate specifically to research must also be applied.”\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes", "responses": [ { "c_id": "4835", "date": "12 Aug 2019", "name": "Owen Driskell", "role": "Author Response", "response": "Thank you for the reviewers helpful comments. In response to the question of DNA. DNA itself - as opposed to the cellular material from which it originates - is not considered to be relevant material under the HT Act. Its storage is therefore not subject to HTA licensing. However, the reviewer is correct. The Act does state to anyone holding cellular material without the qualifying consent of the person/s concerned, intending to analyse the DNA and use the results, may be breaking the law. It is an offence to analyse DNA without qualifying consent unless it is for an excepted purpose. Having human tissue, including hair, nail, and gametes (i.e. cells connected with sexual reproduction), with the intention of its DNA being analysed without the consent of the person from whom the tissue came or of those close to them if they have died is DNA Theft (Medical diagnosis and treatment, criminal investigations, etc., are excluded). Appropriate informed consent is at the heart of Human Tissue Regulation and at the heart of Research Regulation. Pathology professionals must understand their role in research, and the systems in place by which informed consent is achieved, to give confidence to their compliance with this regulation. RNA is not covered by the Human Tissue Act 2004 but the guidance produced by HTA also applies to RNA analysis where it is to be used to provide information about DNA." } ] }, { "id": "50491", "date": "12 Aug 2019", "name": "Bridget Wilkins", "expertise": [ "Reviewer Expertise Governance of biosample-based research within/outwith CTIMPs", "patient involvement with consent for tissue-based research", "biobanking quality standards", "research in the field of haematopathology (specifically", "myeloproliferative neoplasms)." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a valuable brief summary of the regulatory landscape in the UK within which research use of pathology-related biosamples sits. It highlights the core contributions of the national bodies that regulate clinical trials, and the requirements of laboratory accreditation in the UK that underpin quality assurance of pathology samples. The cited sources will provide a useful and regularly updated source of information regarding MHRA, HRA, HTA and UKAS/ISO 15189.\n\nI think that it would be helpful to extend the focus more explicitly beyond the field of clinical trials, to provide confidence for pathologists working in diagnostic settings who wish to undertake research work unrelated to clinical trials. To this end, a little more discussion about how NHS pathology laboratories can achieve HTA licensing for research storage of 'surplus' tissue, the topic of generic/specific consent, and HRA processes for achieving ethical review of research (particularly the processes for research deemed to raise minimal ethical concerns) would all be valuable.\nLegitimate access to a useful dataset of clinical information to accompany tissue samples is often a perceived barrier to pathology-based research, which could usefully be 'myth busted' in an article such as this (HRA and GDPR considerations, in particular).\nIn my opinion, the topics I've touched on above are all inextricably linked to the regulation of tissue-based research and to the encouragement of pathologists to engage in well-regulated research - the 'holy grail' of the NCRI's CM-Path initiative.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes", "responses": [ { "c_id": "4836", "date": "12 Aug 2019", "name": "Owen Driskell", "role": "Author Response", "response": "We thank the reviewer and agree with their comments.The constraints in the scope for this individual article did not permit an exhaustive exploration of the wider details specifically facing NHS laboratories hosting diagnostic archives. We agree wholeheartedly they represent a great resource for potential research that would benefit NHS services and ultimately patients and the care they receive. Further clarity in education and ‘myth-busting’ will benefit the cause of breaking down the perceived barriers to pathology-based research." } ] }, { "id": "52321", "date": "28 Aug 2019", "name": "Tim Helliwell", "expertise": [ "Reviewer Expertise I have spent many years as a clinical academic histopathologist experiencing the importance of appropriate tissue handling when running research projects. In recent years", "as Vice President of the Royal College of Pathologists with responsibility for education and research", "I have been interested in research governance and regulation." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article summarises the main regulatory requirements for the successful involvement of pathologists in research. It will be particularly valuable for diagnostic cellular pathologists who are asked to support clinical trials on an occasional basis, as well as for career researchers with a background in pathology.\nPathologists should be aware that the success or failure of a research project may depend on the quality of material that is analysed; as experts in the handling of tissues and fluids, pathologists are key members of the research team. For human tissues, the conditions of transport and storage (frozen, paraffin wax blocks) may influence the results of research investigations just as much as investigations for clinical practice.\nSpecific comments:\nFigure 1 clearly indicates the potential extent of involvement by pathologists. The accompanying text does not pick up on the 'right hand' green box in the figure which emphasises the opportunity for intellectual input into discussing the results. The scientific understanding provided by pathologists can offer caution in interpreting results, if appropriate, where biological factors e.g. tumour sampling variability, may influence conclusions. A clinician scientists, pathologists are also in a position to discuss the possible clinical impact of the results.\n\nIt might be useful if the authors could state how much of the article is aimed at cellular pathologists and which areas might also be relevant to the other pathology specialities.\n\nIn the section on accreditation, ISO 15189 is more correctly cited as ISO 15189:2012.\n\nAccreditation to the ISO standard implies, as stated in the article, the implementation of an appropriate quality management system. This will include (in the appropriate context) seeking evidence on how a laboratory ensures that research samples are appropriately handled and stored. Factors influencing the uncertainty of measurement are familiar to laboratories, and might impact on the assurance that research sponsors require that samples are of optimal quality for research as well as diagnosis.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-975
https://f1000research.com/articles/8-1654/v1	18 Sep 19	{ "type": "Research Article", "title": "What species make up the Nike fish assemblages at the macrotidal estuary in Gorontalo Bay, Indonesia?", "authors": [ "Femy M. Sahami", "Rene Charles Kepel", "Abdul Hafidz Olii", "Silvester Benny Pratasik", "Rene Charles Kepel", "Abdul Hafidz Olii", "Silvester Benny Pratasik" ], "abstract": "Background: No study has documented the species composition of Nike fish (fam: Gobiidae) schools. The aim of this study is to document the species composition of the Nike-fish schooling. Methods: All samples were collected randomly from fisher’s catch during the fishing season on 5th–11th October 2018 at macrotidal area in Leato. Then, all specimens were identified morphologically by melanophore pattern differences. Subsequently, all identified-samples by melanophores pattern differences were sent to the genetic laboratory for identification. Results: The morphological results show there are five individuals with a different melanophores pattern. On the contrary, the genetic results only show four species from those five individuals. They are Sicyopterus pugnans, S. cynocephalus, Belobranchus segura, and Bunaka gyrinoides. Conclusions: Our findings show that there are only four species that compose the Nike fish schooling in Gorontao Bay. They are Sicyopterus pugnans, Sicyopterus cynocephalus, Belobranchus segura, and Bunaka gyrinoides.", "keywords": [ "Nike-fish", "Gorontalo", "melanophores pattern", "genetic", "morphology" ], "content": "Introduction\n\nEstuaries are a crucial habitat for biota and small fish, in particular juveniles of commercially relevant species. They are considered as the most productive and dynamic ecosystem in the world (Cantera & Blanco, 2001; Lahjie et al., 2019; McHugh, 1967; Sreekanth et al., 2017). They also perform the most crucial role in the population dynamic for a lot of invertebrate and fish species. These ecosystems also significantly contribute to provide some ecological services such as nursery ground, feeding ground and breeding habitats for both freshwater and marine species (Beck et al., 2001; McLusky & Elliott, 2004; Sun et al., 2019). The most well-known species that occupy the seas and estuary area in Gorontalo Bay is Nike fish.\n\nNike (pronounced nee-K) is a local name for transparent juvenile of unknown fish. These fish are approximately 2–4 cm in length; they appear seasonally and fished at estuary waters around the Gorontalo Bay. These juvenile fish has been fished and marketed traditionally for a long time. They are preferable for consumption by the local people than other fisheries products. As a consequence, fishing activity has increased over time to supply local demand for Nike (Wolok et al., 2019).\n\nHowever, the impact of fishing activities is unknown. A recent paper concerning Nike only reports the seasonal appearance during the fishing season (Pasisingi & Abdullah, 2018), total length and morphometric measurements (Zakaria, 2018), nutrition content (Liputo et al., 2013), and mercury contamination of these fish (Salam et al., 2016). To our knowledge, no studies have documented the species diversity that composed the schooling of Nike. Although, Yamasaki et al. (2011) have reported that species in juvenile form can be determined by its melanophores pattern and genetic determination.\n\nThe objective of the present study is to address this lack of knowledge by identifying the fish species that composed a Nike fish schooling. This information is very urgent and required for fisheries management. Therefore, we aimed to identify the species that composed the schooling of Nike fish in Gorontalo Bay by melanophores pattern and genetic identification.\n\n\nMethods\n\nThis study was conducted in October 2018 at Leato (0°30’0.58”N, 123°3’55.42”E), Gorontalo Bay, Indonesia (Figure 2). Approximately 100 g of the Nike-Fish Assemblages (Figure 1) were collected randomly from the fishermen’s catch at fishing grounds during the catch-season (on October 5th–11th). All samples were transported using a cool-box to the lab for measurement. Immediately after collection, all samples were identified visually according to Yamasaki et al. (2011). and the specimens with different melanophore patterns were separated according to their melanophore display. We assummed that those separated individuals were diferent on species.\n\nThe red dot indicates the position of fishing ground where the samples were collected from fishermen.\n\nThen, we selected one individual from each group and labeled these as N1, N2, N3, N4, N5, for genetic identification. Images of the selected samples were captured using Canon EOS 100d with 58 mm pro Digital Wide Converter 0.45X Lens and subsequently converted to black and white using CorelDraw Graphic Suite 2019.\n\nAfter selection, all of the individuals with different melanophores were preserved with ethanol 70% in a separate bottle and sent to the Genetics Laboratory at Manokwari for genetic identification by Sanger sequencing. The DNA cytochrome oxidase subunit I (CO1) of the sample was isolated with a Geneaid™ DNA Isolation Kit. Editing, and proofreadingof seqeunces,, and construction of the the phylogenetic tree was generated with MEGA 5.0 software.\n\n\nResults\n\nFive unspecified individuals of Nike-fish were identified morphologically by melanophore differences, as shown in Figure 3. N1 was revealed as Sicyopterus pugnans; N2 as Sicyopterus cynocephalus; N3 and N5 as Belobranchus segura; and N4 as Bunaka gyrinoides. The specimens with melanophores differences of each group is shown in Figure 4.\n\nNike-fish schools consist of various species with the same body-shape, but different melanophore displays. Moreover, from 100 g (~145 individuals) of the total specimens that we identified, only five individuals with different melanophore patterns were identified (Figure 3).\n\nFigure 3 shows the genetic identification among the individuals (species). The outcomes of genetic identification for N3 and N5 shows that both samples are the same species: Belobranchus segura.\n\n\nDiscussion\n\nAlthough the melanophore patterns in N3 and N5 are different, their genetics are identical, meaning they are the same species (Belobranchus segura). This dissimilarity might be affected by the changes of melanophore during the development of the larvae. Valade et al. (2009) report that such melanophores chang on Sicyopterus langocephalus during the larvae stage. These changes could represent a problem for morphological identification. We can not count the species by morphological differences. Therefore, for the next examination we strongly recommended determining the species composition of the Nike fish schools by genetic rather than morphological identification because for that reason.\n\n\nConclusion\n\nOur findings show that there are four species that compose Nike fish schooling. They are Sicyopterus pugnans, Sicyopterus cynocephalus, Belobranchus segura, and Bunaka gyrinoides.\n\n\nData availability\n\nGroup N1, Sicyopterus pugnans isolate N1_LEATO_1 cytochrome oxidase subunit I (COI) gene, partial cds; mitochondrial. GenBank accession number MN065178.\n\nGroup N2, Sicyopterus cynocephalus isolate N2_LEATO_1 cytochrome oxidase subunit I (COI) gene, partial cds; mitochondrial. GenBank accession number MN069305.\n\nGroup N3, Belobranchus segura isolate N3_LEATO_1 cytochrome oxidase subunit I (COI) gene, partial cds; mitochondrial. GenBank accession number MN069306.\n\nGroup N4, Bunaka gyrinoides isolate N4_LEATO_1 cytochrome oxidase subunit I (COI) gene, partial cds; mitochondrial. GenBank accession number MN069307.\n\nGroup N5, Belobranchus segura isolate N5_LEATO_1 cytochrome oxidase subunit I (COI) gene, partial cds; mitochondrial. GenBank accession number MN069308.", "appendix": "Acknowledgements\n\nThe authors would like to thank La Nane, Sitty Ainsyah Habibie, and Nuralim Pasisingi for technical writing and support during this research.\n\n\nReferences\n\nBeck MW, Heck KL, Able KW, et al.: A better understanding of the habitats that serve as nurseries for marine species and the factors that create site-specific variability in nursery quality will improve conservation and management of these areas. BioScience. 2001; 51: 633–641. Publisher Full Text\n\nCantera JR, Blanco JF: The estuary ecosystem of Buenaventura bay, Colombia. In Coastal marine ecosystems of Latin America. Springer, Berlin, Heidelberg, 2001; 265–280. Publisher Full Text\n\nLahjie AM, Nouval B, Lahjie AA, et al.: Economic valuation from direct use of mangrove forest restoration in Balikpapan Bay, East Kalimantan, Indonesia [version 2; peer review: 2 approved]. F1000Res. 2019; 8: 9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiputo SA, Berhimpon S, Fatimah F: Analisa Nilai Gizi Serta Komponen Asam Amino dan Asam Lemak dari Nugget Ikan Nike (Awaous melanocephalus) DenganPenambahan Tempe. CHEMISTRY PROGRESS. 2013; 6(1). Reference Source\n\nMcHugh JL: Estuarine nekton. In Lauff GH, (Ed.), Estuaries, American Association for the Advancement of Science Special Publication, Washington, DC, 1967; 83: 581–620. Reference Source\n\nMcLusky DS, Elliott M: The Estuarine Ecosystem. Ecology, Threats and Management, Third ed. Oxford University Press, 2004; 214. Publisher Full Text\n\nPasisingi N, Abdullah S: Pola kemunculan ikan Nike (Gobiidae) di Perairan Teluk Gorontalo, Indonesia. DEPIK JurnalIlmu-IlmuPerairan, Pesisir dan Perikanan. 2018; 7(2): 111–118. Reference Source\n\nSalam A, Sahami FM, Panigoro C: Nike (Awaous melanocephalus) Fishery and Mercury Contamination in the Estuary of BoneBolango River. Omni-Akuatika. 2016; 12(2). Publisher Full Text\n\nSreekanth GB, Lekshmi NM, Singh NP: Temporal patterns in fish community structure: environmental perturbations from a well-mixed tropical estuary. Proc Natl Acad Sci India Sect B Biol Sci. 2017; 87(1): 135–145. Publisher Full Text\n\nSun Z, Sokolova E, Brittain JE, et al.: Impact of environmental factors on aquatic biodiversity in roadside stormwater ponds. Sci Rep. 2019; 9(1): 5994. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nValade P, Lord C, Grondin H, et al.: Early life history and description of larval stages of an amphidromous goby, Sicyopterus lagocephalus (Gobioidei: Sicydiinae). Cybium. 2009; 33(4): 309–319. Reference Source\n\nWolok T, Fachrussyah ZC, Yantu I: Technical And Economic Analysis Of Catching Equipment Totaluo In Nike Fishing (Awaous Melanocephalus) In Gorontalo City. Jambura Science of Management. 2019; 1(2): 65–71. Reference Source\n\nYamasaki N, Kondo M, Maeda K, et al.: Reproductive biology of three amphidromous gobies, Sicyopterus japonicus, Awaous melanocephalus, and Stenogobius sp., on Okinawa Island/Biologie de la reproduction de trois gobies amphidromes de l'ile d'Okinawa: Sicyopterus japonicus, Awaous melanocephalus et Stenogobius sp. Cybium, International Journal of Ichthyology. 2011; 35(4): 345–360. Reference Source\n\nZakaria Z: Analisis Morfometrik Schooling Ikan Nike di Perairan Laut Pesisir Kota Gorontalo. Jambura Journal of Educational Chemistry. 2018; 13(1): 77–80. Reference Source" }	[ { "id": "54101", "date": "17 Oct 2019", "name": "Dini Wahyu Kartika Sari", "expertise": [ "Reviewer Expertise Fish Genetics" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nMethods:\nDo not show clearly how many samples of the Nike either for morphological analysis or molecular analysis.\n\nNo information about the size of the Nike.\n\nWhat is the mean of \"The DNA cytochrome oxidase subunit I (CO1) of the sample was isolated with a Genaeid DNA isolation kit\"? It should be genomic DNA.\n\nNo primer information used in this study.\n\nNo information about the PCR mix and the PCR condition.\n\nNo information about how the authors got the sequence result? Sequencing done by who?\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "54104", "date": "23 Oct 2019", "name": "Ken Maeda", "expertise": [ "Reviewer Expertise larval biology of goby" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIf the nike-fish material was composed of Sicyopterus, Bunaka, and Belobranchus species, the larvae must represent different shapes, for example, standard length, head length, preanal length, length of caudal peduncle, fin-ray counts (especially second dorsal, anal and pectoral fins), and fin shapes (especially pelvic fin). Although arrangements of the pigments scattering upon the body surface were shown in the Figure 3, melanophores along the dorsal and ventral midlines are more useful for identification. Morphological identification of these taxa are not difficult at least to the genus level. Please observe the morphologies of the specimens carefully before the molecular identification.\nIntroduction\nSecond paragraph:\nThe first sentence should be “Nike (pronounced nee-k) is a local name for transparent postflexion larvae of fish, but it has not been identified to the species as well as the genus or family level.”\n\nIf you or the local people actually used to know what it is, for example, they are young gobies, please write it.\n\nDoes the “length” mean standard length or total length? Please specify, they are significantly different.\n\nThird paragraph:\nYamasaki et al. (2011) provided key morphological characters (not only the melanophore patterns) to identify species of the newly hatched larvae (not for postflexion larvae and juveniles) of goby. They did not use genetic characters for the larval identification.\n\nMethods\nFirst paragraph:\nHow did you identify the samples visually according to Yamasaki et al. (2011)? They described the morphologies of newly hatched larvae, not the postflexion larvae and juveniles. See the comment above.\n\nWas the collection site the sea, not the estuary? According to the Figure 2, it is marine environment, 150-200 m off from the coast.\n\nSecond paragraph:\nBecause the images have not been used in the manuscript, you don’t need to write the second sentence.\n\nResults\nFirst sentence:\nPlease replace “five unspecified individuals” with “five unspecified types”.\n\nSicyopterus pugnans is a species in Polynesia. So probably it is a misidentification. Please remind that the information in the database is not always correct. Indeed the Sicyopterus pugnans in the figure 4 is divided in to two clades. If they are different species, at least one of them is not the S. pugnans. Please consider the meaning of the results before trusting the information of the database blindly. Please suggest the possibility of misidentification in the Discussion.\n\nMelanophore patterns:\nAs I wrote above, if the nike material was composed of Sicyopterus, Bunaka, and Belobranchus species, the larvae can be identified at least at the genus level by their morphologies. Please observe the specimens carefully before saying “same body shape”.\n\nDiscussion\nI don’t agree with the last two sentences.\n\nFigure 1\nPlease write status of the larvae. Are they living, on ice, or fixed in 70% ethanol?\n\nThe scale bar must be an error. The larvae are too big, if the bar indicates 3 cm. Please confirm.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-1654
https://f1000research.com/articles/8-1653/v1	18 Sep 19	{ "type": "Research Article", "title": "Vaccine hesitancy among parents in Kuala Lumpur: a single center study", "authors": [ "Ahmad Farouk Musa", "Trived Soni", "Xian Pei Cheong", "Rusli Bin Nordin", "Trived Soni", "Xian Pei Cheong", "Rusli Bin Nordin" ], "abstract": "Background: Vaccine hesitancy (VH) is defined as the delay in acceptance or refusal of vaccination despite availability of vaccination services. The main objective of this study was to improve the understanding of vaccine hesitancy (VH) among parents in Kuala Lumpur (KL), Malaysia, by determining the prevalence of VH among parents and to identify the predictors associated with a vaccine hesitant attitude.\n\nMethods: This cross-sectional study was conducted in KL. A questionnaire was devised to collect information from parents, namely sociodemographic information, WHO determinants of VH and the Parents Attitude towards Childhood Vaccine (PACV) scale.\n\nResults: A total of 380 questionnaires were distributed and 337 were returned (88.6% response rate). Those that completed 49 (>90%) out of the 55 given questions in the survey were included for data analysis. Based on inclusion and exclusion criteria, 23 were excluded, giving a sample size of 314. To identify parents, those with at least one child were included, giving a sample size of 221. We noted 60.2% (189) of the participants were females and 80.3% (252) were Malay. Our study found a prevalence of VH of 14.5% among parents based on the 15-item PACV scale. Univariate analysis found no link between sociodemographic factors and VH in parents. Only five of these determinants were included in the final model as statistically significant (p< 0.05) predictors of VH among parents in KL. The five factors were introduction to a new vaccine, negative past experiences of vaccinations, distrust of the pharmaceutical industry, distrust of health systems and providers and being male.\n\nConclusions: Factors contributing towards a prevalence of VH of 14.5% in KL, Malaysia must be studied further to identify any temporal relationship to the under-immunization of children in order to reach the WHO goal of 100% immunization coverage in children and eradication of vaccine preventable diseases.", "keywords": [ "vaccine hesitancy", "parental attitudes to childhood vaccine (PACV)", "prevalence", "determinants" ], "content": "Introduction\n\nHistorically, it is said that Edward Jenner founded vaccine after he successfully inoculated a 13-year old boy with the cowpox virus, who then developed immunity against smallpox1,2. Since then, vaccination has come a long way, with multiple vaccines created for mass production and use.\n\nGlobally, it is found to be the single-most effective intervention to prevent infectious diseases worldwide. Beyond that, vaccines have also been said to prevent antibiotic resistance, empower women, protect against bioterrorism and extend life expectancy3. In 2007, the British Medical Journal found vaccines to be one of four most important developments in medicine in the last 150 years, alongside sanitation, antibiotics and anesthesia4. Unfortunately, we are witnessing a huge rise in vaccine-preventable diseases (VPDs) in Malaysia, where there was an almost 1000% increase in the number of measles cases in 2018 compared to the previous year according to the Ministry of Health (MOH).\n\nThe devastating symptoms and fear of contracting certain diseases resulted in a great uptake of vaccines in their early years, resulting in the eradication of many vaccine-preventable diseases around the world. However, as time passed, more people have become vaccinated and sightings of these diseases have become rare. Thus, public attention has shifted from what was once a fear of a deadly disease to the possible risks and side effects of vaccination. The concerns and questions regarding vaccines gave rise to what we know today as vaccine hesitancy (VH).\n\nVH can be appreciated from as early as the 1800s, after the first vaccine was created by Edward Jenner. People complained about the invasiveness of methods used in administering vaccines. In addition to that, scientists questioned the basis for its efficacy, while religious groups rejected vaccination due to its unnatural ways and as an act of ‘playing God’5.\n\nThe epitome of vaccine controversy occurred at the end of the 19tth century, when Dr Andrew Wakefield published an article in The Lancet regarding the positive association between the measles, mumps and rubella (MMR) vaccine and bowel disease and autism6. This paper was formally retracted in view of Wakefield’s conflict of interest, as well as committing scientific fraud by falsifying data. However, the influence of this report still lingers in many societies.\n\nThe World Health Organization (WHO) recently estimated that 19.4 million infants are missing out on basic vaccines, while one in five children worldwide do not receive routine life-saving vaccines7. Similarly, the United Nations International Children's Emergency Fund (UNICEF) has estimated that 2.7 million children die every year due to VPDs, of which 40% are from the Asian continent8. Furthermore, we are witnessing a similar trend in previously eradicated diseases, such as diphtheria and pertussis, which are once again infecting children in Malaysia.\n\nThe term ‘determinant of vaccination’ includes the many dimensions and expressions of VH. The WHO’s Strategic Advisory Group of Experts (SAGE) recognized the term ‘determinants’ as covering the topic of barriers and enablers of vaccination. After a review of models and discussion among experts regarding this topic, this working group recognized three categories of determinants of VH (Figure 1)9.\n\nReprinted from Vaccine; 32(19); Heidi J. Larson, Caitlin Jarrett, Elisabeth Eckersberger, David M.D. Smith, Pauline Paterson; Understanding vaccine hesitancy around vaccines and vaccination from a global perspective: A systematic review of published literature, 2007–2012; 2150–2159; Copyright (2014);25 with permission from Elsevier.\n\nContextual influences consist of historic, environmental, economic, political, social, cultural and institutional factors that might influence VH populations10. Historically, the most famous incident is the relationship proposed between MMR and autism by Andrew Wakefield, published in The Lancet. The article gained interest and spread on many media sites, resulting in a decrease in MMR vaccination acceptance worldwide. Despite retracting this paper and disproving the theory in multiple follow up studies, parents still report concerns on safety of these vaccines11. In France, the national vaccination schedule was suspended in the 1990’s due to a reported association between hepatitis B vaccination and multiple sclerosis12–15. Despite evidence disapproving the association, this belief remained in the community showing VH in small clusters of general practitioners16.\n\nUnintentional suspicion may arise, especially when national policies are applicable to all and freedom of choice is not given to the public17. This distrust in the authorities and concerns regarding vaccine safety are sometimes shared among physicians as well16. Certain groups are commonly anti-vaccination, such as hard-liner naturopaths, professionals such as some groups of chiropractors, and some religious sects. The spreading of this belief by these groups can influence people to develop VH18–20. Some orthodox protestants believe that it is an act against God and any side effects are the consequence of Gods punishment21. Muslims require any consumed product to be Halal, and if vaccines are not branded Halal, they may be reluctant or refuse it altogether22,23.\n\nIn addition, the modern culture is such that the public questions the validity of science, especially when not backed by research evidence. Patient autonomy makes individuals easily influenced against vaccination, as they act based on their own knowledge and experiences in the past, which may not be updated or just be a misconception24.\n\nIndividual or group influences are influences arising from personal perception of the vaccines or influences of the social and peer environment10. The issue of vaccine safety is the most common determinant of VH. Vaccines are believed to be unsafe in terms of both short-term adverse reactions and long-term side effects. Nearly all the literature observes this reason for non-vaccination among VH groups9,25–28. Personal beliefs about the benefits of vaccines also influence vaccine uptake, with 67.6% of HPV non-vaccinees in the United States stating they consider it not at all or only somewhat beneficial29. Parents were found to be concerned about overwhelming their child’s immune system and the risk of developing the disease after vaccination30. Interestingly, some physicians have doubts regarding the usefulness of certain vaccines themselves, and were therefore not recommending them; meningitis C and hepatitis B are examples31,32.\n\nThe social norm has been found to be a positive reason for vaccine uptake33. Brown noted that having peers or relatives that were pro-MMR vaccine influenced vaccine uptake in the respondents34. In addition, viewing vaccination as a responsibility to society in terms of achieving herd immunity saw a higher vaccine uptake in a systematic review by Quadri-Sheriff35.\n\nRisk perceptions influence vaccination decision36. From a lay person’s point of view, risk is based on past experiences rather than evidenced-based medicine37,38. Besides that, weighing the risk of experiencing side effects with the risk of contracting the already rare VPDs plays a role in VH39. This is known as ‘omission bias’. The need for these vaccines then becomes questionable, as most commonly seen in human papillomavirus (HPV) vaccinations. Parents believe that their children are too young and presume that they are not sexually active; hence, do not require the vaccine40,41. Elderly females, on the other hand, claim that they are too old or no longer sexually active and don’t need the vaccine30,42.\n\nVaccines or vaccine-specific issues are directly or indirectly related to the VH attitude10. Every vaccine has its own production process and problems associated with it. Hence, it is subject to different degrees of hesitancy. There have been doubts about the reliability and constituents of specific vaccines such as thimerosal, albumin or aluminum-based additives43.\n\nMost vaccines are expensive to produce, considering the amount of skilled lab work required. A survey found that half of pediatricians and family physicians surveyed in the United States had delayed purchase of specific vaccines in their practice due to cost (49%) and experienced decreased profit margins for these vaccines (53%)44,45. Fewer healthcare facilities providing vaccines may translate into poorer access for the public and hence, lower vaccination coverage. A similar problem is faced by patients, with higher prices being a barrier to vaccination. Those with private medical insurance show higher vaccination coverage than those without insurance43,46.\n\nAnother common reason for parents to not vaccinate their children is inconsistent advice from physicians. Interestingly, approximately 1 in 5 pediatricians dismiss families who refused one or more vaccines during their consultation. O’Leary discovered that these behaviors by physicians was related to the absence of philosophical exemption law and having a medium or difficult policy for attaining an exemption47. These physicians may feel overwhelmed by the concerns expressed by patients and it may jeopardize the doctor-patient relationship. This is alarming, as it is not only encouraging VH but repudiates access to healthcare for these patients. Others have found the complexity of the vaccine schedule and frequent updates to it very confusing. This then becomes a barrier, as physicians are not confident or comfortable discussing vaccination with parents48,49. The same applies to patients, who find the regimes too complicated to adhere to50.\n\nThe aim of this study was to improve our understanding of VH among parents in Malaysia, specifically looking at the Kuala Lumpur region. The objectives are as follows:\n\n1. To determine the prevalence of VH among parents in Kuala Lumpur, Malaysia.\n\n2. To identify predictors associated with a VH attitude in parents in Kuala Lumpur, Malaysia.\n\n\nMethods\n\nA cross-sectional research design was used for the purpose of this study, where sociodemographic and WHO determinants of VH were studied in relation to VH behavior in parents. This was the most suitable design, as it allowed us to demonstrate the relationship between multiple variables to the outcome at a point in time. The study was conducted from December 2016 to May 2017.\n\nEthics clearance was obtained from the Monash University Human Research Ethics Committee (MUHREC: 9216) and the Medical Research Ethics Committee (MREC: NMRR-16-2508-33624[IIR]) of the Malaysian Ministry of Health before the commencement of the study.\n\nTo calculate the required sample size for a single proportion, we used Pocock’s formula: n= Zα2P (1-P)/d2, where n = minimum required sample size, (Zα) = 1.96, d (precision) = 5%, and P = expected prevalence. Based on our extensive literature search, the prevalence of VH was found to be within the range of 10% to 30%. Using these percentages, respectively, an estimated sample size of between 138 and 318 was required. The average gave the required sample size of 227. Considering potential dropouts and incomplete forms, 227 + (20%) = 275 was our target sample size. A total of 380 questionnaires were distributed.\n\nThe specific inclusion criteria for a parent to participate in this study were as follows: aged 20 years or older; provided written consent; able to understand and comprehend English or Malay languages. The exclusion criteria were being too sick or in an uneasy state to complete questionnaire and being non-citizens of Malaysia.\n\nConvenience sampling was adopted for recruiting patients from Tanglin Health Community Clinic, Kuala Lumpur. Patients were approached by the on-site investigator in the waiting area of the Maternity Clinic at Tanglin Health. All conversations and forms were given in either Malay or English language, depending on the preference of the participant. They were first asked if they would like to participate in a survey regarding VH. Patients that were agreeable were then given a patient information sheet51 while the investigator briefly explained the purpose of the study and their role as participants. Eligible participants were given a separate consent form51 to obtain their signature and name as written consent. Once they had consented, they were handed the self-administered questionnaire, which was normally completed within 10–15 minutes and returned to the investigator.\n\nThe questionnaire consisted of questions regarding socio-demographic characteristics and WHO determinants of VH (a copy of the questionnaire is available as Extended data)51. While this questionnaire was not piloted or validated, it was based on SAGE Working Group on Vaccine Hesitancy’s matrix of determinants of vaccine hesitancy and the PACV scale, which have been validated52,53. The matrix basically mapped out the key factors that influenced the decision to either accept, delay, or reject vaccination altogether. Three major domains were explored; namely, contextual, individual and group, and vaccine-specific.\n\nAll data computation was performed using the Statistical Package for Social Sciences (SPSS) version 23.0, provided by Monash University Malaysia. The data was initially checked for normality using the stem and leaf plot and the Kolmogorov-Smirnov test (p>0.05 suggests normally distributed data).\n\nSociodemographic findings from the returned questionnaire were described using frequencies and percentages. For the prevalence and predictors of VH among parents, only participants who were parents were studied. This was determined using the inclusion criteria of having at least one child.\n\nFindings from the second part (WHO determinants of VH), that look into the three main domains were depicted as frequencies and percentages of ‘YES’ and ‘NO’ responses, except for questions one, two and three. All thirty-five questions were reviewed by a clinician in the field to determine right (non-hesitant) and wrong (hesitant) answers to score and ease analysis. From question four to 35, all ‘NO’ responses were hesitant responses, except for questions four, five, 11, 14, 21, 26, 30, 32, 33 and 35. For question one (source of information for vaccine) and two (reference when hearing negative comment about vaccine), participants were allowed to choose more than one answer. Overall, the response was portrayed in a bar chart. Responses for question one were subclassified into two or more sources (non-hesitant) compared to anything less than two sources (hesitant). As for question two, those who used a doctor as reference (non-hesitant) were compared to reference/consultation of all other groups (hesitant). For question three (which group that are anti-vaccine advocates are the most influential), participants were allowed to choose one answer only. Those that picked any of the choices were scored as hesitant, while the remainder who responded ‘None’ were scored as non-hesitant. All hesitant answers were scored as one and non-hesitant were scored as zero. For subdomains, the score of the questions were added up and the mean was used as the cut-off mark, where anything more than the mean was scored as one and anything less was scored as zero.\n\nThe outcome of VH in parents was determined by the Parent Attitudes about Childhood Vaccines (PACV) scale. The data obtained from the 221 parents were represented in a table using frequencies and percentages. The method and scoring used was according to the guidelines obtained from Opel26. The score given for each item was two for hesitant responses, one for unsure responses and zero for low/no hesitancy. For questions one and two, any ‘Don’t Know’ (DK) responses were excluded as missing data in the analysis. The total score from the scale was divided by the maximum score (30) to calculate the percentage. Any score above 50% was classified as a vaccine hesitate parent (VHP), while scores below that were classified as no/low hesitancy to vaccines. In the case of missing values due to ‘DK’ responses in questions one and two, the score was divided against the maximum adjusted score, where one ‘DK’ response = 28 max score and two ‘DK’ responses = 26 max score. Overall, the final scores for all respondents were used to determine the prevalence of VH among parents in percentage.\n\nSociodemographic factors and WHO determinants of VH were then examined for association with VH using chi square tests (Pearson Chi Square test, Fisher Exact test, and Mantel-Haenszel test for trend) with statistical significance set at p<0.05. Univariate binary logistic analysis for these variables was then carried out using odd ratios (ORs) and associated 95% confidence intervals (95% CIs) to determine significant predictors of VH among parents. From the analysis, all significant predictors (p<0.05) were included in multivariate binary logistic regression using the stepwise method (backward logistic regression). The gave the final model, showing significant and independent predictors of VH in parents.\n\n\nResults\n\nA total of 380 questionnaires were distributed and 337 were returned (88.6% response rate). Those that completed 49 (>90%) out of the 55 given questions in the survey were included for further data analysis. For this reason, 22 participants were excluded, which gave a sample size of 314. To identify parents for the VH in parents analysis, parents with at least one child were included, giving a sample size of 221.\n\nA total of 314 questionnaires were included in the demographic data analysis51. Of these, 250 and 64 completed the Malay and English versions of the questionnaire, respectively. The ages of participants ranged from 20 to 80 years old. The most common age group was ‘30–39’ years followed by ‘20–29’. Figure 2 demonstrates the distribution of ages of the participants as a histogram. The average age of collected respondents were 39.25 ± 13.033 years.\n\nAs shown in Table 1, in terms of gender, 60.2% (189) of the participants were female. There was a predominance of Malays 80.3% (252) compared to all other ethnic groups 19.7% (62). These other groups consisted of 31 Indian, 23 Chinese and eight other ethnicities. Likewise, Islam was the most common religion at 81.8% (257). Non-Muslims (28.2%, 57) comprised of 25 Hindus, 18 Christians, 13 Buddhists, and one other.\n\nThe most common household income was the ‘RM 2,000 - RM 5,000’ category, reported by 143 respondents (47.5%). Household incomes of ‘< RM 2,000’ and ‘RM 5,000 - RM 10,000’ were reported by approximately 20% of the respondents for both groups.\n\nIn terms of education, there was a decreasing trend, with the highest proportion of participants completing secondary education (39.1%), followed by a diploma (34.3%). Only six participants had achieved the highest level of education; a doctorate.\n\nFigure 3 shows the number of children of each participant. The average number of children for all the respondents were two children (mean=1.91) and 88 participants had no children. Parents who had at least one child made up 221 of our respondents. This was further sub-divided into those with ‘one or two children’ and ‘more than two children’, with 119 and 102 respondents in these groups, respectively.\n\nFinally, perceived immunization status showed that approximately 70% have had their children fully or partially immunized, while the remaining the 30% reported that their children had not been immunized at all. This perception was also seen in terms of self-vaccination status, in regard to HPV, influenza, the diphtheria, tetanus and pertussis booster and varicella vaccines.\n\nAs shown in Figure 4, the majority of the respondents used television and the internet as their primary source of information regarding vaccines, reported by 50% and 52.2% of participants, respectively. Other utilized sources were newspapers (34.1%) and social networking sites (25.4%) such as Facebook. The use of radio or magazines as a source of information regarding vaccines was less common, with magazines being the least (11.5%).\n\nAs illustrated in Figure 5, upon receiving negative information regarding vaccination, the majority of participants (75.8%) acted by consulting doctors. The majority of the remaining participants referred to sources on the internet (27.1%). A small number of participants consulted their colleagues or spouses about concerns related to unfavorable statements regarding vaccines. The group consulting religious leaders was reported by only 5.4% of participants.\n\nFigure 6 shows that more than half (61.3%) of respondents claimed to be unaware of any groups against vaccines. Amongst the known anti-vaccine groups, complementary/alternative medicine practitioners (21.1%) were found to be most influential, followed closely by religious groups/leaders (11.5%). On the other hand, traditional medicine and political group/leaders were less influential in exerting impact regarding anti-vaccination.\n\nReferring to Table 2, a small number of participants (13.4%) reported having negative experiences related to vaccination, which discourages them from permitting vaccination for their own children. We noticed that 23.2% of respondents reported having known or being acquainted with individuals who are hesitant towards vaccines due to religious beliefs. When asked their opinion, approximately 14.4% held the opinion that parents who refused vaccination for their children were not putting their life or their children’s life at danger. From the questionnaire, the majority (92.6%) expressed their assurance in the capabilities of the Ministry of Health (MOH) in making decisions regarding vaccinations. A similar number of participants agreed that vaccination should be made compulsory. Likewise, 12.3% of respondents would forgo vaccination for their children if it required a commute of more than an hour. A lower but still a sizable percentage of participants (83.0%) trusted the safety and effectiveness of vaccines produced by pharmaceutical companies.\n\nAll hesitant answers are ‘NO’ except for questions four and five.\n\nReferring to Table 3, 41 participants (13.1%) claimed to know of people who had experienced severe side effects following vaccination. Roughly two thirds (65.5%) of the respondents denied knowing of cases in which unvaccinated children had suffered from illness because they were not vaccinated. Only 23 participants (7.4%) held the belief that vaccines do not boost the body’s immune response. Alarmingly, a large proportion of participants (38.9%) believed that there are superior alternatives to vaccines in preventing illness. When asked about their knowledge of vaccines, 40.8% admitted they were ignorant on which vaccines were suitable for them or their children. In fact, about half of the participants (49.5%) had not heard of the HPV vaccination prior to taking the survey. A small proportion, but not an insignificant one, of 12.7% believed that polio vaccines were redundant and no longer required.\n\nAll hesitant answers are ‘NO’ except for questions 11, 14 and 21.\n\nAs shown in Table 4, most participants believed that vaccines are safe for them or their children. However, 45.6% of participants responded that they did not receive sufficient information regarding vaccines and their safety and 55.7% indicated their reluctance in letting their child be the first recipient of a new vaccine. With regard to the method of administering the vaccine, 40.5% of participants indicated a preference for methods other than injection. The process of vaccination was found to be simple and easy with most participants being confident that the clinics/hospital they visit are well equipped with the vaccines they require. A total of 57.6% of participants are unaware of the MOH’S vaccination schedule and 37.5% of respondents believed that some vaccines were difficult to obtain due to the vaccine schedules. Despite that, the vast majority (93.5%) believed that vaccines should be taken according to schedules. Nearly all participants believed vaccines had a significant value despite being administered to children for free. However, when asked about their willingness to bear the cost of vaccines, 34.1% replied negatively. Furthermore, 12.6% of participants were reluctant on returning to a healthcare provider due to mistreatment and 26.9% claimed to have received advice from a healthcare provider that a certain vaccine was not required.\n\nAll hesitant answers are ‘NO’ except for questions 26, 30, 32, 34 and 35.\n\nOur study shows a prevalence of VH of 14.5% (32/221 patients) among parents, based on the 15-item Parental Attitudes to Childhood Vaccine (PACV) scale (Table 5). Hesitancy was calculated based on Opel26 scoring (see Methods).\n\nTable 6 demonstrates the responses received from 221 parents on the PACV scale. A total of 18.1% of parents have postponed vaccination due to reasons other than illness or allergy. Of these, only 9.0% didn’t get these vaccines at all. Most parents were confident of the MOH childhood vaccination schedule. However, nearly half of the respondents thought that children get too many vaccinations than are needed and were concerned about the safety and side effects of vaccines.\n\nAt least a third of parents agreed that it is better for children to develop natural immunity by getting sick rather than being vaccinated and it is better for children to receive fewer vaccine shots simultaneously. We also noted that 30.3% of parents were concerned about the efficacy of vaccines in preventing illness. Only 2.7% responded that if they had another infant, they would not want them to receive all the recommended vaccinations and 3.6% responded that they were hesitant about vaccinations. Furthermore, only 12.7% of parents were not confident in their child’s doctor.\n\nTable 7 shows the findings of univariate binary logistic regression of all the factors studied in regard to VHP. In terms of sociodemographic characteristics, it was found that male parents are most likely to be VH as compared to female parents (OR 2.431, p = 0.022, 95%CI 1.135-5.204). Other factors such as age, income, education status and number of children did not show any association to hesitancy in our study.\n\nB, coefficient for the constant; S.E., standard error around the coefficient for the constant; Wald, Wald chi-square test; Sig, significance (significant if p<0.05); df, degrees of freedom; Exp(B): exponentiation of the B coefficient.\n\nFor contextual influence, there was no association for the number of sources of information or consulting anyone other than a doctor when hearing a negative comment about vaccines. Similarly, being aware of any influential group/leaders that advocate for anti-vaccination and being influenced by historical events were not associated with hesitancy among parents. Trust in MOH regarding health policies was found to be a highly significant factor, with those answering negatively being ten-times more likely to be hesitant (OR 10.23, p < 0.001, 95%CI 3.464-30.197). There was also an association with VH and negative responses to questions regarding trust in pharmaceutical industries (OR 10.29, p < 0.001, 95%CI 4.322-24.476) and geographical barriers to vaccination (OR 4.42, p = 0.002, 95%CI 1.736-11.251).\n\nGiving a negative responses to WHO individual/group influence questions was found to be associated with hesitant behavior among parents.\n\nIn terms of vaccine/vaccination specific issues, being apprehensive towards new vaccines (OR 8.387, p – 0.001, 95%CI 2.452-28.686) or the design of vaccines and the procedure (OR 10.81, p < 0.001, 95%CI 3.827-30.547) resulted in being approximately 10 times more likely to be VH. Having issues with vaccination schedule (OR 2.684, p = 0.017, 95%CI 1.190-6.051), cost of vaccines (OR 3.347, p = 0.003, 95%CI 1.511-7.418), scientific evidence regarding the risk and benefit of vaccines (OR 18.26, p < 0.001, 95%CI 4.412-75.580) was also associated with VH.\n\nA multivariate (stepwise) binary logistic regression analysis was undertaken to assess the relationship between demographic and WHO determinants as predictors of VH among parents. Following stepwise regression using the forward conditional model, the full model containing the five remaining predictors (male gender, distrust of the pharmaceutical industry, distrust of the health system/providers, negative past experiences of vaccinations and reluctance to be the first recipient for new vaccines) were statistically significant in predicting the incidence of VH. The model explained between 29.8% (Cox & Snell R square) and 53.0% (Nagelkerke R square) of the variance in VH among parents and correctly classified 85.6% of cases.\n\nAs shown in Table 8, the five independent variables made a unique statistically significant contribution to the model. The strongest predictor of VH was reluctance to be the first recipient for new vaccines (OR: 25.00, p = 0.001, 95% CI 3.571-166.67), followed by negative past experiences of vaccinations (OR: 8.214, p = 0.004, 95% CI 1.991-33.879), distrust of the health system/providers (OR: 6.173, p = 0.003, 95% CI 1.835-20.833), being a male parent (OR: 4.608, p = 0.009, 95% CI: 1.477-14.493) and distrust of the pharmaceutical industry (OR: 2.431, p < 0.001, 95% CI 1.135-5.204).\n\nB, coefficient for the constant; S.E., standard error around the coefficient for the constant; Wald, Wald chi-square test; Sig, significance (significant if p<0.05); df, degrees of freedom; Exp(B): exponentiation of the B coefficient.\n\n\nDiscussion\n\nOur study discovered a prevalence of VHPs of 14.5%. This was higher than the figure of 11.6% as discovered by Azizi and collaborators54. This rate might translate to a poorer uptake of vaccines in the future55.\n\nIn Malaysia, Ezat discovered that 12.9% of respondents did not accept HPV vaccination for their daughters. However, this study was conducted among mothers only, rather than parents of both genders56. In a recent survey, the National Health and Morbidity Survey 201657, it was revealed that only 86.4% of children in Malaysia has complete immunization coverage. This is still considered unsatisfactory. Overall, these studies show a trend of VH behavior among 10-15% of parents in Malaysia.\n\nHowever, Eve Dubé has recorded a high prevalence rate of 30% in her multi-country study on VH58. Although the prevalence levels found here in Malaysia are lower, we cannot tell if the problem will continue growing rampantly. Necessary actions should be taken to curb this behavior before it becomes a bigger problem for outbreaks of VPDs in the future. It has been said that “vaccination is a victim of its own success”17, considering the fact that vaccination programs have been so successful and that VPDs are becoming less visible; therefore, people are now focused on the risk and alleged risk of vaccines, rather than the risk of contracting the diseases they protect us from.\n\nIntroduction to a new vaccine. Introduction to a new vaccine was the strongest predictor of VH among parents. More than half would not agree to be the first to vaccinate their child when a new vaccine was released. In the past, when the HPV vaccination was newly released, there was poor uptake of the vaccine as it was done on a voluntary basis, rather than being subsidized and compulsory as part of the national immunization schedule. Logistically, new vaccines are more expensive to manufacture and hence, the element of cost may be a reason for this poor uptake.\n\nSafety and efficacy might be another reason for this, with at least 40–50% of parents reporting concern in regard to this. This same phenomenon was observed when the H1NI vaccine was introduced in 2009. There was a very low uptake rate of H1NI vaccine, which was mainly attributed to concerns about the safety of what was perceived to be a newly developed drug59. We see a similar trend in Malaysia. Despite having approximately two million cases of people infected with dengue per year60, the introduction of a new vaccine, which was approved in Mexico in 2015, is still stalled due to safety fears. The idea of waiting for the perfect vaccine concoction with high efficacy and satisfactory safety profile may be one of the reasons for VHPs. Unless a new vaccine is proven to be safe in the long-term, this factor remains to be the most important in determining the confidence of parents in vaccines.\n\nNegative past experiences. While the quality of vaccination services might have a positive influence on vaccine acceptance, negative past experiences of vaccination or vaccination services can also affect the acceptance of vaccination. Nowak and Cacciatore61 have shown that personal experience of bad or adverse events following vaccination correlates with the lower confidence levels. They have also shown that confidence levels were higher in parents who did not have a child who experienced any bad reactions or did not know of anyone who had. Similarly, stories about vaccine adverse events, including those reported by the media, would affect the confidence levels of parents regarding vaccination.\n\nWe therefore would suggest that special attention should be given to this finding. Medical practitioners and healthcare workers should always be aware that any experience of the parents and also those indirectly acquired from others might influence their confidence level. They perhaps should interview the parents in order to learn about their experiences and then make an extra effort to explain the misconceptions and provide the information on how rarely these complications occur.\n\nIt has been shown that anti-vaccine activists have been spreading fears about vaccination, especially about the safety of vaccines. The fear about the MMR vaccine casing autism is still being spread widely in our community and this secondary ‘experience’ might have an impact on parents. Although there is little data on the correlation between anti-vaccine websites and decision-making by parents regarding vaccination, it is alarming to know that studies have shown that parents who delayed or refused vaccines are most likely to be those who have sought knowledge regarding vaccines on the internet62.\n\nTrust in the pharmaceutical industry. The main reason for the lack of trust in the pharmaceutical industry may be due to the social stigma that pharmaceutical companies are profit-driven and do not care about the safety and efficacy of their products. We have seen an all-time low level of trust in large pharmaceutical companies that manufacture and promote vaccines. This fear of what is known as ‘pharmaceutical-industrial complex’ has led some parents to believe that there is a conspiracy in the vaccine industry63. These parents argued that the results of vaccination research are influenced by pharmaceutical lobbying influences64.\n\nOther possible reasons include the use of foreign products such as thimerosal, albumin or aluminum-based additives. Being unsure of the reliability and possible harmful effects of these constituents may play a role in the VH of parents27. In addition, porcine constituents are one of the major concerns for many, as it is against the ‘Halal’ principle in the Islam religion, followed the majority of our participants65. There is a misguided belief that vaccines contain porcine DNA, hence making them impure and forbidden for use by Muslims54.\n\nTrust in the health system and provider. The term ‘trust’ has an array of meanings in the medical literature. To put it simply in the context of vaccine and vaccination, it is about making a risk versus benefit decision when someone has incomplete or inadequate information66. A few studies have shown that VH is closely related to the idea of trust in health professionals and health providers, including the health system and health institutions67,68. Trust in the health system and health providers, who recommend the use of vaccines and determine the vaccination schedules, has a heavy bearing on the development of VH among parents59.\n\nBenin et al.20 noticed that trust, or the lack of it, forms the basis for parents to either vaccinate or not vaccinate their children, while Larson et al. have pointed out that trust placed in the healthcare system, in the healthcare workers, and also in the policy makers who decide on the vaccination programs, mediates the impact of these factors on vaccine hesitancy17. It is also interesting to note that the level of trust in the healthcare system is much lower among religious and ethnic minorities66. Perhaps this distrust could be traced back to way these minority groups were either neglected or discriminated against by the government that provides the healthcare system66. This is a possible explanation to the finding by Azizi et al.54 that non-Muslims in Malaysia were more vaccine hesitant than Muslims, since they were in the minority of the population in this country and felt they were being discriminated against.\n\nThis is a determinant which is not discussed widely in the literatures. However, Siddiqui et al.69 found that there was noticeable gender differences among parents. In their study, women were shown to have more trust than men in vaccines. We have also seen that research on barriers to vaccination rarely records the views of men and families related to vaccination70. However, women and social and familial status play a very crucial role in facilitating the process of vaccination.\n\nIt is almost a global phenomenon that men or fathers are rarely involved in vaccination programs by the government. Therefore, information regarding vaccines or VPDs normally does not reach them71. By targeting only women, vaccination interventions somehow neglect the influence of men as head of the family in Asian culture in general, and in Malaysian culture in particular. Due to this lack of exposure to information on vaccines and VPDs, men may be more susceptible to the kinds of influences discussed above and subsequently have a bearing on women’s decisions about vaccination as well. This, to us, is the only probable cause for this VH among men due to being neglected in education programs regarding vaccination.\n\nOur study is one of the first few studies to explore the WHO determinants and VH among parents. In addition to the recent literature in this field, our study allows for a better understanding of VH behavior in Malaysia. The findings of this study are hoped to help transparency and awareness, which is important in gaining trust and guiding parents’ decisions on childhood vaccination. Pharmaceutical companies should work together with government healthcare bodies for the primary prevention of VH behavior in parents. For example, this can be done through health campaigns or health talks for parents in schools, or even through mass media for easier accessibility. Doubts, concerns, misconceptions and new healthcare policies and vaccines should be addressed publicly. Information sources should be made available on online platforms, as an easy reference for parents.\n\nParental decisions are mainly influenced by doctors. Hence it is crucial that doctors can answer patients’ questions regarding vaccines. When faced with doubts or concerns, they should be well equipped with sufficient evidence from multiple sources to justify the need for vaccines in children and provide relevant information upon the request of parents. Regular Continuous Medical Education (CME) sessions should be given to keep doctors up to date on recent literature on vaccines. MOH healthcare facilities should encourage and seek feedback from patients, including about confidence and trust in MOH itself. An audit should be done annually to identify any specific issues or concerns, which should be attended to in an appropriate and timely manner.\n\nThere are certain limitations to our study to be recognized. First, the study design precludes the possibility of causality between the factors tested with VH. The findings from the study are also based on the self-perception of the parent at one particular time. This could potentially change over time, place and context. As the participants were given a choice of whether to answer each question, there was several missing values which could not be accounted for in our study. It is unsure if this is due to a lack of understanding of any of the questions or any other specific reasons. The study could not account for other potentially significant predictors of VH because the independent risk factors in the model could only explain 29.8%% to 53.0%% of the variance in VH. Further studies should explore other factors that may be related to VHPs in a multi-ethnic country, such as an association of cultural or social differences with VH in parents. A qualitative research design may be necessary to explore this gap of knowledge in our diverse community setting.\n\nSelection bias was also seen, as participants were included only if they could read and understand Malay or English Language. This may affect the findings as it excluded other languages spoken in Malaysia, such as Mandarin and Tamil. When looking at number of children, mothers pregnant with their first child and their spouses were excluded, as this section was filled with zero. Using convenience sampling to recruit our patients resulted in a highly skewed ethnic and religious population, where nearly 90% consisted of Malays and Muslims. This is not reflective of the overall Malaysian population; hence, the findings cannot be generalized to all Malaysian parents. Ideally, these factors should be taken into consideration using a simple random sampling method to recruit participants to achieve a more accurate population that can be generalized to all Malaysian parents.\n\nFinally, this study is based in a single center in an urban area of Kuala Lumpur. The responses and the results might be different in a semi-urban or rural population.\n\n\nConclusion\n\nTo summarize, the prevalence of VH is reported at 14.5% in parents in Kuala Lumpur, Malaysia. The main determinants described by WHO SAGE working group were studied in this population and five factors were identified as significant predictors of VH: namely, reluctant to be the first recipient of a new vaccine, negative past experiences, distrust in the pharmaceutical industry, distrust in the health system and provider, and of male gender. These factors must be recognized and studied further to identify any temporal relationship to under-immunization in children. Equipped with this, hopefully we can play our part in reaching the WHO goal of 100% immunization coverage in children and the eradication of VPDs.\n\n\nData availability\n\nHarvard Dataverse: Replication Data for: Vaccine hesitancy among parents in Kuala Lumpur: a single center study. https://doi.org/10.7910/DVN/BU5HGV51.\n\nThis project contains the following underlying data:\n\nVaccine Hesitancy Raw Data.tab (demographic information and questionnaire responses for each participant)\n\nVaccine Hesitancy OUTPUT.pdf (raw statistical analysis output data)\n\nHarvard Dataverse: Replication Data for: Vaccine hesitancy among parents in Kuala Lumpur: a single center study. https://doi.org/10.7910/DVN/BU5HGV51.\n\nThis project contains the following extended data:\n\nPIS & Consent Form (ENGLISH).pdf\n\nPIS & Consent Form (MALAY).pdf\n\nQuestionnaire Proforma - Vaccine Hesitancy (ENGLISH).pdf\n\nQuestionnaire Proforma - Vaccine Hesitancy (MALAY).pdf\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "References\n\nThe History of Vaccines: The College of Physicians of Philadelphia. Reference Source\n\nYaqub O, Castle-Clarke S, Sevdalis N, et al.: Attitudes to vaccination: A critical review. Soc Sci Med. 2014; 112: 1–11. PubMed Abstract \| Publisher Full Text\n\nAndre FE, Booy R, Bock HL, et al.: Vaccination greatly reduces disease, disability, death and inequity worldwide. Bull World Health Organ. 2008; 86(2): 140–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFerriman A: BMJ readers choose the “sanitary revolution” as greatest medical advance since 1840. BMJ. 2007; 334(7585): 111. Publisher Full Text \| Free Full Text\n\nJacobson RM, St Sauver JL, Finney Rutten LJ: Vaccine hesitancy. Mayo Clin Proc. (CONCISE REVIEW FOR CLINICIANS) (Report). 2015; 90(11): 1562–1568. PubMed Abstract \| Publisher Full Text\n\nHackett AJ: Risk, its perception and the media: the MMR controversy. Community Pract. (PROFESSIONAL). 2008; 81(7): 22–5. PubMed Abstract\n\nImmunization coverage. World Health Organization; 2017. Reference Source\n\nThe State Of Asia-Pacific's Children. Nairobi, Kenya: UNICEF; 2008. Reference Source\n\nMacDonald NE; SAGE Working Group on Vaccine Hesitancy: Vaccine hesitancy: Definition, scope and determinants. Vaccine. 2015; 33(34): 4161–4. PubMed Abstract \| Publisher Full Text\n\nHeidi Larson EK: Rapid literature review on motivating hesitant population groups in Europe to vaccinate. 1 ed. Stockholm: ECDC — European Centre for Disease Prevention and Control; 2015.\n\nStefanoff P, Mamelund SE, Robinson M, et al.: Tracking parental attitudes on vaccination across European countries: The Vaccine Safety, Attitudes, Training and Communication Project (VACSATC). Vaccine. 2010; 28(35): 5731–7. PubMed Abstract \| Publisher Full Text\n\nSadovnick AD, Scheifele DW: School-based hepatitis B vaccination programme and adolescent multiple sclerosis. Lancet. 2000; 355(9203): 549–50. PubMed Abstract \| Publisher Full Text\n\nMonteyne P, André FE: Is there a causal link between hepatitis B vaccination and multiple sclerosis? Vaccine. 2000; 18(19): 1994–2001. PubMed Abstract \| Publisher Full Text\n\nAscherio A, Zhang SM, Hernán MA, et al.: Hepatitis B vaccination and the risk of multiple sclerosis. N Engl J Med. 2001; 344(5): 327–32. PubMed Abstract \| Publisher Full Text\n\nMikaeloff Y, Caridade G, Rossier M, et al.: Hepatitis B vaccination and the risk of childhood-onset multiple sclerosis. Arch Pediatr Adolesc Med. 2007; 161(12): 1176–82. PubMed Abstract \| Publisher Full Text\n\nVerger P, Fressard L, Collange F, et al.: Vaccine Hesitancy Among General Practitioners and Its Determinants During Controversies: A National Cross-sectional Survey in France. EBioMedicine. 2015; 2(8): 891–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLarson HJ, Cooper LZ, Eskola J, et al.: Addressing the vaccine confidence gap. Lancet. 2011; 378(9790): 526–35. PubMed Abstract \| Publisher Full Text\n\nSalmon DA, Haber M, Gangarosa EJ, et al.: Health consequences of religious and philosophical exemptions from immunization laws: individual and societal risk of measles. JAMA. 1999; 282(1): 47–53. PubMed Abstract \| Publisher Full Text\n\nO'Shea T: Vaccination Is Not Immunization. 4th ed. 2015; 210. Reference Source\n\nBenin AL, Wisler-Scher DJ, Colson E, et al.: Qualitative analysis of mothers' decision-making about vaccines for Infants: The Importance of Trust. Pediatrics. 2006; 117(5): 1532–41. PubMed Abstract \| Publisher Full Text\n\nRuijs WL, Hautvast JL, van Ijzendoorn G, et al.: How orthodox protestant parents decide on the vaccination of their children: a qualitative study. BMC Public Health. 2012; 12(1): 408. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWoo YL, Razali SM, Chong KR, et al.: Does the success of a school-based HPV vaccine programme depend on teachers' knowledge and religion? - A survey in a multicultural society. Asian Pac J Cancer Prev. 2012; 13(9): 4651–4. PubMed Abstract \| Publisher Full Text\n\nNordin SA, Idris F, Hamat RA, et al.: Parental refusal to diphtheria vaccine: A fatal outcome. Med J Malays. 2013; 68(5): 435–6. PubMed Abstract\n\nKane MA: Commentary: public perception and the safety of immunization. Vaccine. 1998; 16 Suppl: S73–5. PubMed Abstract \| Publisher Full Text\n\nLarson HJ, Jarrett C, Eckersberger E, et al.: Understanding vaccine hesitancy around vaccines and vaccination from a global perspective: A systematic review of published literature, 2007-2012. Vaccine. 2014; 32(19): 2150–9. PubMed Abstract \| Publisher Full Text\n\nOpel DJ, Taylor JA, Mangione-Smith R, et al.: Validity and reliability of a survey to identify vaccine-hesitant parents. Vaccine. 2011; 29(38): 6598–605. PubMed Abstract \| Publisher Full Text\n\nHarmsen IA, Mollema L, Ruiter RA, et al.: Why parents refuse childhood vaccination: a qualitative study using online focus groups. BMC Public Health. 2013; 13: 1183. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nButler R, MacDonald NE; SAGE Working Group on Vaccine Hesitancy: Diagnosing the determinants of vaccine hesitancy in specific subgroups: The Guide to Tailoring Immunization Programmes (TIP). Vaccine. 2015; 33(34): 4176–9. PubMed Abstract \| Publisher Full Text\n\nRosenthal SL, Weiss TW, Zimet GD, et al.: Predictors of HPV vaccine uptake among women aged 19–26: Importance of a physician's recommendation. Vaccine. 2011; 29(5): 890–5. PubMed Abstract \| Publisher Full Text\n\nLyren A, Leonard E: Vaccine refusal: Issues for the primary care physician. Clin Pediatr (Phila). 2006; 45(5): 399–404. PubMed Abstract \| Publisher Full Text\n\nKarafillakis E, Dinca I, Apfel F, et al.: Vaccine hesitancy among healthcare workers in Europe: A qualitative study. Vaccine. 2016; 34(41): 5013–5020. PubMed Abstract \| Publisher Full Text\n\nDubé E, Laberge C, Guay M, et al.: Vaccine hesitancy: An overview. Hum Vaccines Immunother. 2013; 9(8): 1763–73. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNichter M: Vaccinations in the Third World: a consideration of community demand. Soc Sci Med. 1995; 41(5): 617–632. PubMed Abstract \| Publisher Full Text\n\nBrown K, Fraser G, Ramsay M, et al.: Attitudinal and demographic predictors of measles-mumps-rubella vaccine (MMR) uptake during the UK catch-up campaign 2008-09: cross-sectional survey. PLoS One. 2011; 6(5): e19381. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nQuadri-Sheriff M, Hendrix KS, Downs SM, et al.: The role of herd immunity in parents' decision to vaccinate children: a systematic review. Pediatrics. 2012; 130(3): 522–530. PubMed Abstract \| Publisher Full Text\n\nBrewer NT, Chapman GB, Gibbons FX, et al.: Meta-analysis of the relationship between risk perception and health behavior: the example of vaccination. Health Psychol. 2007; 26(2): 136–145. PubMed Abstract \| Publisher Full Text\n\nHobson-West P: Understanding vaccination resistance: moving beyond risk. Health Risk Soc. 2003; 5(3): 273–283. Publisher Full Text\n\nCasiday RE: Children's health and the social theory of risk: insights from the British measles, mumps and rubella (MMR) controversy. Soc Sci Med. 2007; 65(5): 1059–1070. PubMed Abstract \| Publisher Full Text\n\nSerpell L, Green J: Parental decision-making in childhood vaccination. Vaccine. 2006; 24(19): 4041–4046. PubMed Abstract \| Publisher Full Text\n\nWamai RG, Ayissi CA, Oduwo GO, et al.: Awareness, knowledge and beliefs about HPV, cervical cancer and HPV vaccines among nurses in Cameroon: an exploratory study. Int J Nurs Stud. 2013; 50(10): 1399–1406. PubMed Abstract \| Publisher Full Text\n\nLaz TH, Rahman M, Berenson AB: An update on human papillomavirus vaccine uptake among 11-17 year old girls in the United States: National Health Interview Survey, 2010. Vaccine. 2012; 30(24): 3534–40. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLaz TH, Rahman M, Berenson AB: Human papillomavirus vaccine uptake among 18- to 26-year-old women in the United States: National Health Interview Survey, 2010. Cancer. 2013; 119(7): 1386–92. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMyth and Realities: Responding to arguments against vaccination. 2013; 5: 30. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFreed GL, Cowan AE, Clark SJ: Primary care physician perspectives on reimbursement for childhood immunizations. Pediatrics. 2008; 122(6): 1319–24. PubMed Abstract \| Publisher Full Text\n\nFreed GL, Clark SJ, Cowan AE, et al.: Primary care physician perspectives on providing adult vaccines. Vaccine. 2011; 29(9): 1850–4. PubMed Abstract \| Publisher Full Text\n\nNeubrand TP, Breitkopf CR, Rupp R, et al.: Factors Associated With Completion of the Human Papillomavirus Vaccine Series. Clin Pediatr (Phila). 2009; 48(9): 966–9. PubMed Abstract \| Publisher Full Text\n\nO'Leary ST, Allison MA, Fisher A, et al.: Characteristics of physicians who dismiss families for refusing vaccines. Pediatrics. 2015; 136(6): 1103–11. PubMed Abstract \| Publisher Full Text\n\nBrownlie J, Howson A: ‘Between the demands of truth and government’: health practitioners, trust and immunisation work. Soc Sci Med. 2006; 62(2): 433–43. PubMed Abstract \| Publisher Full Text\n\nPetousis-Harris H, Goodyear-Smith F, Turner N, et al.: Family physician perspectives on barriers to childhood immunisation. Vaccine. 2004; 22(17-18): 2340–4. PubMed Abstract \| Publisher Full Text\n\nGoad J, Durham M: The anti-vaccine movement: A pharmacist's view. Vaccinophobia and Vaccine Controversies of the 21st Century. Springer New York; 2013; 119–28. Publisher Full Text\n\nMusa AF: \"Replication Data for: Vaccine hesitancy among parents in Kuala Lumpur: a single center study\". Harvard Dataverse, V1, UNF: 6: 8Nf0E+ixnR7ksxjkZhP/DQ== [fileUNF]. 2019. http://www.doi.org/10.7910/DVN/BU5HGV\n\nLarson HJ, Jarrett C, Schulz WS, et al.: Measuring vaccine hesitancy: The development of a survey tool. Vaccine. 2015; 33(34): 4165–75. PubMed Abstract \| Publisher Full Text\n\nOpel DJ, Magione-Smith R, Taylor JA, et al.: Development of a survey to identify vaccine-hesitant parents: the parent attitudes about childhood vaccines survey. Hum Vaccin. 2011; 7(4): 419–25. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMohd Azizi FS, Kew Y, Moy FM: Vaccine hesitancy among parents in a multi-ethnic country, Malaysia. Vaccine. 2017; 35(22): 2955–61. PubMed Abstract \| Publisher Full Text\n\nOpel DJ, Taylor JA, Zhou C, et al.: The relationship between parent attitudes about childhood vaccines survey scores and future child immunization status: A validation study. JAMA Pediatrics. 2013; 167(11): 1065–71. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEzat SW, Hod R, Mustafa J, et al.: National HPV immunisation programme: knowledge and acceptance of mothers attending an obstetrics clinic at a teaching hospital, Kuala Lumpur. Asian Pac J Cancer Prev. 2013; 14(5): 2991–9. PubMed Abstract \| Publisher Full Text\n\nLim KK, Chan YY, NoorAni A, et al.: Complete immunization coverage and its determinants among children in Malaysia: findings from the National Health and Morbidity Survey (NHMS) 2016. Public Health. 2017; 153: 52–57. PubMed Abstract \| Publisher Full Text\n\nDubé E, Gagnon D, Nickels E, et al.: Mapping vaccine hesitancy--country-specific characteristics of a global phenomenon. Vaccine. 2014; 32(49): 6649–54. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRamanadhan S, Galarce E, Xuan Z, et al.: Addressing the Vaccine Hesitancy Continuum: An Audience Segmentation Analysis of American Adults Who Did Not Receive the 2009 H1N1 Vaccine. Vaccines. 2015; 3(3): 556–78. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWoon YL, Hor CP, Lee KY, et al.: Estimating dengue incidence and hospitalization in Malaysia, 2001 to 2013. BMC Public Health. 2018; 18(1): 946. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNowak GJ, Cacciatore MA: Parents' confidence in recommended childhood vaccinations: Extending the assessment, expanding the context. Hum Vaccin Immunother. 2017; 13(3): 687–700. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSmith PJ, Humiston SG, Marcuse EK, et al.: Parental delay or refusal of vaccine doses, childhood vaccination coverage at 24 months of age, and the Health Belief Model. Public Health Rep. 2011; 126 Suppl 2: 135–46. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPoland GA, Jacobson RM, Ovsyannikova IG: Trends affecting the future of vaccine development and delivery: the role of demographics, regulatory science, the anti-vaccine movement, and vaccinomics. Vaccine. 2009; 27(25–26): 3240–44. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDubé E, Vivion M, Sauvageau C, et al.: “Nature Does Things Well, Why Should We Interfere?”: Vaccine Hesitancy Among Mothers. Qual Health Res. 2016; 26(3): 411–25. PubMed Abstract \| Publisher Full Text\n\nJeyachelvi K, Juwita S, Norwati D: Human Papillomavirus Infection and its Vaccines: Knowledge and Attitudes of Primary Health Clinic Nurses in Kelantan, Malaysia. Asian Pac J Cancer Prev. 2016; 17(8): 3983–8. PubMed Abstract\n\nLarson HJ, Clarke RM, Jarett C, et al.: Measuring trust in vaccination: A systematic review. Hum Vaccin Immunother. 2018; 14(7): 1599–09. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStreefland P, Chowdhury AMR, Ramos-Jimenez P: Patterns of vaccination acceptance. Soc Sci Med. 1999; 49(12): 1705–16. PubMed Abstract \| Publisher Full Text\n\nHobson-West P: ‘Trusting blindly can be the biggest risk of all’: organised resistance to childhood vaccination in the UK. Sociol Health Illn. 2007; 29(2): 198–15. PubMed Abstract \| Publisher Full Text\n\nSiddiqui M, Salmon DA, Omer SB: Epidemiology of vaccine hesitancy in the United States. Hum Vaccin Immunother. 2013; (12): 2043–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCoreil J, Losikoff P, Pincu R, et al.: Cultural feasibility studies in preparation for clinical trials to reduce maternal-infant HIV transmission in Haiti. AIDS Educ Prev. 1998; 10(1): 46–62. PubMed Abstract\n\nSen G, Östlin P, George A: Unequal, Unfair, Ineffective and Inefficient. Gender Inequity in Health: Why it exists and how we can change it. Final report to the WHO Commission on Social Determinants of Health, WHO; 2007. Reference Source" }	[ { "id": "54109", "date": "02 Jun 2020", "name": "Italo Francesco Angelillo", "expertise": [ "Reviewer Expertise Public Health" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe aims of the study were to understand vaccine hesitancy among parents in Kuala Lumpur, Malaysia, by determining the prevalence of VH among parents and to identify the predictors associated with a vaccine hesitant attitude. Substantial revisions are necessary to be addressed. The paper also needs editing by a native English person.\n\nIntroduction\nThe introduction is too long and should be considerably condensed.\nMethods\nIt is not indicated how many Health Community Clinic are present in the geographic area. Please clarify.\n\nIt is not given any information regarding a pilot study for testing the survey questionnaire. Please clarify.\n\nIt should be indicated who and how many investigators approached the sample. Please clarify.\n\nThe Authors should clarify about the face-validity testing of the questions with an explanation of the validity of the content of the questions with regard to the research aims.\n\nThe Authors should clarify if the questionnaire was anonymous and no personally identifiable information was collected.\n\nThe Authors should clarify whether the participants have received any gift or monetarily compensated.\n\nOne of the major methodological weakness is that the sample size calculation indicated that 275 was the target size and a total of 380 questionnaires were distributed. However, when parents with at least one child were included the giving sample size was only 221. Since one of the aims of the study was to determine the prevalence of Vaccine hesitancy among parents, it is not clear why the selection was not restricted only to parents.\n\nThe statistical analysis is not, strictly speaking, adequate, because it would be particularly relevant to describe the model developed, the variables included, and the rationale why they are included. Moreover, it should be indicated if the statistical tests were one-side or two-sides. It would be particularly relevant to include in the model the source of information.\nResults\nNo information is given about how many individuals refused to participate and the response rate should be included. If not all patients have agreed to participate, no information is given about them. Was there any attempt to quantify the response bias: information about non-responders. It would be useful to have some kind of indication of comparability with non-respondents. Is there any population-based data available? How did they differ from those in the sample, how representative is the sample and were the findings representative?\n\nThe characteristics of the sample should be considerably condensed.\n\nThe results of the ORs and of the 95% CI should be with only two decimals.\nDiscussion\nIt should be inserted a deeper comment regarding the results of previous surveys conducted in different geographic areas on vaccine hesitancy. The following articles should be cited: Dubè Human Vaccines & Immunotherapeutics 2019;15(1):113-1201. Napolitano et al. Human Vaccines & Immunotherapeutics 2018;14:1558-1565)2.\n\nPage 16: It is stated that Trust in the health system and health providers, who recommend the use of vaccines and determine the vaccination schedules, has a heavy bearing on the development of VH among parents. The role of health professionals is vital and additional references regarding other studies in order to support the statement should be added. Cite: D’Alessandro et al. Hum Vaccin Immunother 2018;14:1573-93; Ragan et al. Vaccine 2018;36:331-414; Napolitano et al. Hum Vaccin Immunother 2016;12:1504-105.\n\nThe paragraph regarding the main limitations of the study should discuss all limits such as, for example, the recall bias and the social desirability bias.\nFigures and Tables\nTables 7 and 8 should be condensed and should report only the OR (with only two decimals), the S.E., the 95% CI (with only two decimals), and the p-value.\n\nFigures 1 to 6 should be deleted.\n\nTable 5 should be deleted.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "64443", "date": "26 Aug 2020", "name": "Regina Célia de Menezes Succi", "expertise": [ "Reviewer Expertise Pediatric Infectious Diseases", "Vaccines", "HIV" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript from Ahmad Farouk Musa and colleagues has the objective of improving the understanding of vaccine hesitancy among parents in Kuala Lumpur, by determining the prevalence of VH and identify predictors associated with this attitude. The results are interesting to this region, but some questions should be addressed.\nThe Introduction contains more information than needed. It should be shorter and more objective.\n\nThere are too much Tables and Figures with similar data. All the figures should be excluded.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "64442", "date": "26 Aug 2020", "name": "Anju Aggarwal", "expertise": [ "Reviewer Expertise child neurology and immunization" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a good study.\n\nIntroduction seems to be too long as if a combination of review and a study.\n\nIntroduction does not need a illustration from another study or book.\n\nSample size has been calculated and ethical clearance has been obtained. The article can be indexed after modification.\n\nA separate review article could be created.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1653
https://f1000research.com/articles/8-646/v1	09 May 19	{ "type": "Software Tool Article", "title": "Network-based Observability and Controllability Analysis of Dynamical Systems: the NOCAD toolbox", "authors": [ "Dániel Leitold", "Ágnes Vathy-Fogarassy", "János Abonyi", "Dániel Leitold", "Ágnes Vathy-Fogarassy" ], "abstract": "Network science has become increasingly important in life science over the last decade. The proposed Octave and MATLAB-compatible NOCAD toolbox provides a set of methods which enables the structural controllability and observability analysis of dynamical systems. In this paper, the functionality of the toolbox is presented, and the implemented functions demonstrated.", "keywords": [ "Dynamical systems", "Complex networks", "Controllability and observability analysis", "Robustness", "MATLAB toolbox" ], "content": "Introduction\n\nIn the life sciences, the determination of driver nodes in networks that play a significant role in the emergence or treatment of diseases is an intensively researched field1. In large-scale human liver metabolic networks (HLMN), the driver metabolites have essential functions, and the role of transport reactions and extracellular metabolites in terms of controlling HLMN has revealed the importance of the environment of human liver metabolism with regard to the health of the liver2.\n\nIn terms of controlling the human signalling network, the role of different proteins was also systematically analysed with the toolset of network controlability in 3 to highlight the role of cancer-associated genes. Target control with objective-guided optimisation (TCO) was introduced to control a set of variables (or targets) of interest while the quantity of drivers and constrained nodes were minimised and maximised, respectively. This method is capable of determining the leading phenotype transitions in biological networks that can be identified as drug targets4. Using statistical analysis, a subset of critical control nonprotein-coding RNAs (ncRNAs) enriched by human disease can also be determined5. In intra-cellular networks, to understand the information flow, a natural control system was utilised and the robustness of such a control was analysed6. The importance of determining the proper driver nodes in biological networks, or more generally in any dynamical system, is unequivocal, and the amount of research concerning network science has increased rapidly. A detailed study about the control principles in biological networks has already been published7. The network science-based analysis of dynamical systems has spread rapidly as it provides simple and efficient tools to analyse the structural controllability of any linear or linearised system1.\n\nAlthough considerable research has utilised the method8, a flexible software tool which may be used to support the research in this field has yet to be designed. Parallel research has resulted in a collection of applications, toolboxes, plug-ins and scripts that analyse and determine several structural properties of genes, protein-protein interaction or even social or urban networks. Most of these applications only analyse the structural properties of static networks and just a handful of them utilise these structural properties to draw conclusions concerning the dynamics of the system investigated. As our toolbox belongs to the second group, in the following section, the available applications and programs of this group are elaborated.\n\nA brief summary of the available tools with expanded functionalities is given in Table 1. Applications or software packages implemented in Python and capable of analysing the controllability and observability of dynamical systems are: graph-control9 and WDNfinder10. The advantage of Python-based development lies in its widespread use and the countless methods and packages implemented in this language, including the tools developed for network analysis11. Although in Python the focus is on developing a broad software package for complex system analysis, this has yet to be fulfilled and all of the available solutions have limitations. The graph-control toolbox only analyses the impact of network topology on the number of inputs and implements the fast matching algorithm12. Even though WDNFinder only determines the minimum driver node set (MDS) and classifies nodes based on MDS, it is incapable of facilitating extended analysis.\n\nAdditionally, the CytoCtrlAnalyser13 plug-in for Cytoscape14 has been developed, which was implemented in Java and offers graphical interfaces for users as well. It evaluates control centrality, control capacity and classifies nodes for biomolecular networks. Furthermore, the Ecological Network Analysis with R software package (enaR) provides some dynamical analysis functions and can generate models to analyse ecological networks in the R environment15. As can be seen, both software packages deal with special kinds of networks. The netctrl program can determine the driver nodes and switchboard dynamics model for any complex network16. CONTEST is a MATLAB toolbox which can analyse the dynamics of complex systems, but these dynamics do not cover the structural controllability and observability properties17 of the analysed system. Although the presented software packages ensure the design of a controllable and observable system, they do not provide the opportunity to analyse the designed system exhaustively. These functions are helpful in terms of supporting the work of experts, but are insufficient for the sophisticated analysis of systems.\n\nThe contribution of this paper is to provide a novel toolbox, NOCAD18, for the comprehensive analysis of linear or linearised dynamical systems based on the approach of network science. In the following section, the implemented functions and measurements are presented through examples of their application.\n\n\nMethods\n\nWith the help of the presented Ocatave- and MATLAB-compatible toolbox, experts can create, analyse and improve any type of dynamical systems. As the structure of the dynamical systems is generally represented by their adjacency matrix and linear dynamical systems can be described by the state-space model that contains the state-transition, input, output and feedthrough matrices, the Octave/MATLAB programming language is a perfect environment to handle these matrices and provide comprehensive functionalities based on them. With the use of NOCAD18, experts and researchers can effectively determine the input and output matrices of state-space models, calculate system-specific qualitative measurements (e.g. diameter, relative degree, control centrality and robustness of the system, etc.) and improve the system to satisfy the relative degree-based requirements. The workflow of the toolbox can be seen in Figure 1.\n\nThe network mapping module provides two methods to create a dynamical system based on the topology of the state variables. The system characterisation module generates more than 49 measures to analyse, classify and characterise the developed system. The improvement and robustness module offers five algorithms to improve the system with additional inputs (observers) as well as outputs (controllers), and can analyse the robustness of the designed system.\n\nThe functions of the toolbox can be performed step-by-step given its modular structure. Each module has a specific task and one function from each module calls the others. A system can be analysed by calling the main functions from the modules. The advantage of this structure is its modularity as each module can be expanded easily and further modules also implemented in a simple way. A list of their functions and dependencies on each other is presented in the manual.\n\nAccording to the aforementioned approach, the implemented functions of the toolbox were divided into three modules as follows: (1) network mapping module, (2) system characterisation module and (3) improvements and robustness module.\n\nThe network mapping module creates a dynamical system from a given network structure, i.e. the necessary matrices of the state-space model are generated for the topology in such a way, that the created system is structurally controllable and structurally observable. The determination of the input and output matrices can be achieved by the path finding and signal sharing methods19, which modify the result of the maximum matching algorithm.\n\nThe system characterisation module performs the calculation of 49 numerical measures to qualify the dynamical system based on its structure. The implemented measures, on the one hand, are well-known static measures (e.g. the number of nodes and edges, closeness and betweenness centralities), and, on the other hand, measures that characterise the dynamics of the system (e.g. structural controllability, observability, control centrality and relative degree). This module can also be used for the purpose of simple network analysis.\n\nThe improvement and robustness module integrates two main functions. On the one hand, it enables the input and output configurations of the system to be extended in such a way that the relative degree of the modified system does not exceed the initially defined threshold. For this purpose, this module implements five methods, namely the set covering-based grassroot and retrofit methods20, the centrality measures-based method20, the modified Clustering Large Applications based on Simulated Annealing algorithm (mCLASA), and the Geodesic Distance-based Fuzzy c-Medoid Clustering with Simulated Annealing algorithm (GDFCMSA)20,21. On the other hand, this module allows users to examine the robustness of the extended configurations by removing nodes from the network representation and by checking the structural controllability and structural observability of the damaged system.\n\nThe implemented methods are introduced in detail in the cited articles and the manual of the NOCAD toolbox.\n\nIn order to use the NOCAD toolbox18, installation of Octave or MATLAB is required. Then the directories of the toolbox must be copied into the working directory, or the directories of the toolbox must be added to the paths. The functions were implemented in Octave 5.1.0 and MATLAB R2016a on a Windows 64-bit system. On other operating systems, or with other Octave or MATLAB versions, proper operation is not guaranteed. Our toolbox is independent of other MathWorks toolboxes, it uses only the octave-networks-toolbox22 and the greedy set covering implementation23.\n\n\nUse cases\n\nIn this section, the main functionalities of the NOCAD toolbox18 presented through examples of use cases. Although many biological networks are available from public databases, due to their complex nature, they are unsuitable for such a simple illustration. Therefore, the services of the NOCAD toolbox are presented on simple artificial networks.\n\nThe first step in each workflow is to create a state-space model from the adjacency matrix that presents the structural description of the system. This can be achieved by the use of path finding and signal sharing methods implemented in the first module. Both methods are modified versions of the maximum matching algorithm. An example of the application of the path finding method for the creation of a state-space model from the adjacency matrix (A) is shown in Figure 2. In this figure, B denotes the resulting input matrix, C the output matrix, while D stands for representing the direct feedthrough.\n\nThe network represents the A state transition matrix. B denotes the input matrix in which the places of the nonzero elements are determined by the controller node allocation algorithm. Similarly, the C output matrix is defined with the observability analysis of the network of the state variables. The D matrix of the direct feedthrough contains only zeros.\n\nAs the configuration above is not complex enough to demonstrate the functions of the second module, a more complex configuration of the input and output nodes is used. The sample input and output configurations can be seen in Figure 3, where the input and the output nodes are denoted by blue and red, respectively.\n\nThe system presented in Figure 3 consists of 9 state variables and 15 directed connections between them. Quality measures calculated by the System characterisation module of the NOCAD toolbox can be seen in Figure 4, Figure 5, and Figure 6.\n\nIn Figure 4, measures qualifying the whole system with one value are presented. The density shows that the number of edges is almost a fifth of the possible maximum, and the diameter of the system (i.e. the longest shortest path in the network that presents its structure) is 4. The degree variance is 2.67, while the Freeman’s centrality is 0.43. The relative degree of the system is also 4. The Pearson coefficient shows that the in-in and in-out correlations are assortative in nature, while out-out and out-in correlations are likely to be disassortative. The system is controllable and observable. As no loop is present in the network, the percentage of loops relative to edges is 0%. As there are 6 edges that have symmetric edge pairs and the number of connections is 15, the percentage of the symmetric edge pairs relative to the edges is 40%.\n\nNode centrality measures assigned to the state variables of the system are also presented in Figure 4. One of the most important values is the highest degree of the nodes, which belongs to state variable x4. As Scott’s centrality is a normalised degree, the most important node is once again x4. The closeness of node xi is calculated as the ratio of the number of nodes reachable from xi to the sum of their distances from xi. The higher value indicates the more central position of the node, and, once again, node x4 is the most central element. The betweenness centrality shows how many shortest paths intercept the given node. If a node has a high value, then it is a critical node in the structure. The highest value belongs to nodes x2 and x4. The PageRank assigns a percentage value for each node, based on their centrality roles if Markov-chains are modelled. The measure referred to as correlation shows the proportion of the number of edges of neighbours’ and the number of neighbours. This information is useful when determining the assortativity of the system. The control centrality and observe centrality measures determine how many state variables can be influenced or observed by the nodes.\n\nIn Figure 5, the first vectors (referred to as driver and sensor nodes) show the driver and sensor nodes as logical vectors. The following four vectors classify these nodes as source, external, internal and inaccessible driver and sensor nodes. These types of nodes are introduced in 24 in detail. In the next section of the figure, the controlling and observing matrices are presented. Generally, these matrices are sparse matrices, as only the columns of drivers and sensors contain nonzero values. In Figure 5, we converted them into row vectors for their appropriate visualisation. The values show the number of derivations necessary to influence or observe a state variable in the system. Next, the similarity of the driver and sensor nodes is presented. The similarity of driver nodes x4 and x6 is 0.81. In this case, the reason why it is less than 1 is that although they control the same set of nodes, the numbers of derivations that influence them are different. In terms of sensor similarity, sensor nodes x2 and x3 observe the same set of nodes and they do this almost simultaneously, so their similarity is 0.91. R𝒞 and R𝒪 are the simple reachability matrices. They show which nodes can be controlled or observed by a given node. In R𝒞, the ith column shows which nodes can control node i. From the other viewpoint, elements in row i highlight those nodes which can be controlled by node i. In this example, node x8 can influence every node, but it does not guarantee structural controllability. The R𝒪 matrix can be interpreted analogously with regard to observability.\n\nFinally, measures of edge centrality are seen in Figure 6. The betweenness has the same meaning as in the case of nodes, that is, it yields the number of shortest paths that intercept the edge25. From this perspective, the most critical edge is the edge a46 with a value of 10. The endpoint similarity shows how similar the influenced and observed sets of the state variables with regard to the endpoints of edges are. This metric has a high value if the edge is part of a cycle or creates a bridge in the network. As no bridges are present in this network, only cycles can be recognised by this measure. The edge similarity shows how similar the roles of edges are, and it allows redundancies, to be located. In the topology presented, nodes x1, x2 and x3, or nodes x4, x5, x6 and x7 also create parts of the network that possess redundancy.\n\nFor the demonstration of the last module, configurations provided by the first module are used again (Figure 2). Results provided by this module can be seen in Figure 7. In this case, five methods were applied to the system to extend the configuration as follows: the required relative degree was set at 2, while the alpha parameter of the cost function was set at 0.521. Results show that all the methods determine the same set of driver nodes for the system, that is, they are sufficient to influence state variables x4 and x8. The resultant cost is 1.5556, the relative degree is 2 which satisfies the requirements, and the mean of the relative degrees is 1.1111. In this configuration, six different nodes can be identified which can be damaged separately and the system remains controllable. This is expressed by the value of robustness (66.6%). The most important nodes in terms of controllability are x2, x4 and x7. In the case of observability, methods yield different solutions with the exception of the centrality measures-based and mCLASA algorithms which provide the best configuration in this case. Although the cost as well as the maximum and mean of the relative degree were identical in the case of retrofit set covering-based and GDFCMSA methods as well, the robustness analysis of these configurations exhibits a higher degree of vulnerability.\n\n\nConclusions\n\nIn this article the Octave- and MATLAB-compatible NOCAD toolbox18 was proposed to support the network-based controllability and observability analysis of dynamical systems. The toolbox offers two methods to design a structurally controllable and observable system based on the state-transition matrix. The designed system can be analysed by 49 qualitative measures both from structural and dynamical points of view. The toolbox serves five methods to improve the designed system by adding new inputs and outputs to it, thus, its relative degree can be decreased. Then the robustness of the individual designs can also be evaluated. The modular structure of the toolbox supports the facile improvement of the modules by adding new functions and the toolbox can be extended by new modules as well. Even though the modules are built on each other, most of their functions can also be used independently from each other.\n\nAlthough our goal in this paper is to draw the attention of researchers of life sciences to the services provided by the NOCAD toolbox, it can be utilised in practice in various fields of sciences as well, for example, it enables social networks to be controlled in the economy, transaction networks to be analysed in finance or dynamical systems to be designed in engineering.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\n\nSoftware availability\n\nSource code available from: https://github.com/abonyilab/NOCAD.\n\nArchived source code at time of publication: https://doi.org/10.5281/zenodo.265667418\n\nLicense: GNU General Public License v3.0\n\n\nAuthor contributions\n\nDániel Leitold reviewed the literature on network science, developed the algorithms, implemented the Octave and MATLAB functions, designed as well as performed the experiments, and wrote the related sections. Ágnes Vathy-Fogarassy participated in the formalisation of the methodology. János Abonyi developed the algorithms, implemented the Octave and MATLAB functions and proofread the paper.", "appendix": "Grant information\n\nThis research was supported by the National Research, Development and Innovation Office NKFIH, through the project OTKA-116674 (Process mining and deep learning in the natural sciences and process development) and the EFOP-3.6.1- 16-2016- 00015 Smart Specialization Strategy (S3) Comprehensive Institutional Development Program. Dániel Leitold was supported by the ÚNKP-18-3 New National Excellence Program of the Ministry of Human Capacities.\n\nThe funders had no role in study design, data collection and analysis and decision to publish.\n\n\nReferences\n\nLiu YY, Slotine JJ, Barabási AL: Controllability of complex networks. Nature. 2011; 473(7346): 167–73. PubMed Abstract \| Publisher Full Text\n\nLiu X, Pan L: Detection of driver metabolites in the human liver metabolic network using structural controllability analysis. BMC Syst Biol. 2014; 8(1): 51. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiu X, Pan L: Identifying driver nodes in the human signaling network using structural controllability analysis. IEEE/ACM Trans Comput Biol Bioinform. 2015; 12(2): 467–72. PubMed Abstract \| Publisher Full Text\n\nGuo WF, Zhang SW, Shi QQ, et al.: A novel algorithm for finding optimal driver nodes to target control complex networks and its applications for drug targets identification. BMC Genomics. 2018; 19(Suppl 1): 924. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNacher JC, Akutsu T: Controllability methods for identifying associations between critical control ncrnas and human diseases. Methods Mol Biol. In Computational Biology of Non-Coding RNA. 2019; 1912: 289–300. PubMed Abstract \| Publisher Full Text\n\nRavindran V, Nacher JC, Akutsu T, et al.: Network controllability analysis of intracellular signalling reveals viruses are actively controlling molecular systems. Sci Rep. 2019; 9(1): 2066. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLi M, Gao H, Wang J, et al.: Control principles for complex biological networks. Brief Bioinform. 2018. PubMed Abstract \| Publisher Full Text\n\nLiu YY, Barabási AL: Control principles of complex systems. Rev Mod Phys. 2016; 88(3): 035006. Publisher Full Text\n\nChaturvedi V: Controllability of networks. Reference Source\n\nChu Y, Wang Z, Wang R, et al.: Wdnfinder: A method for minimum driver node set detection and analysis in directed and weighted biological network. J Bioinform Comput Biol. 2017; 15(5): 1750021. PubMed Abstract \| Publisher Full Text\n\nZinoviev D: Recognize-Construct-Visualize-Analyze-Interpret. Pragmatic Bookshelf. 2018. Reference Source\n\nFaradonbeh MKS, Tewari A, Michailidis G: Optimality of fast-matching algorithms for random networks with applications to structural controllability. IEEE Trans Control Netw Syst. 2017; 4(4): 770–780. Publisher Full Text\n\nWu L, Li M, Wang J, et al.: Cytoctrlanalyser: a cytoscape app for biomolecular network controllability analysis. Bioinformatics. 2018; 34(8): 1428–1430. PubMed Abstract \| Publisher Full Text\n\nShannon P, Markiel A, Ozier O, et al.: Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003; 13(11): 2498–504. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBorrett SR, Lau MK: enaR: an R package for ecosystem network analysis. Methods Ecol Evol. 2014; 5(11): 1206–1213. Publisher Full Text\n\nNepusz T, Vicsek T: Controlling edge dynamics in complex networks. Nat Phys. 2012; 8(7): 568–573. Publisher Full Text\n\nTaylor A, Higham DJ: Contest: A controllable test matrix toolbox for matlab. ACM Trans Math Softw. 2009; 35(4): 26. Publisher Full Text\n\nAbonyi J: abonyilab/nocad v2.0. 2019. http://www.doi.org/10.5281/zenodo.2656674\n\nLeitold D, Vathy-Fogarassy Á, Abonyi J: Controllability and observability in complex networks–the effect of connection types. Sci Rep. 2017; 7(1): 151. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeitold D, Vathy-Fogarassy Á, Abonyi J: Evaluation of the complexity, controllability and observability of heat exchanger networks based on structural analysis of network representations. Energies. 2019; 12(3): 513. Publisher Full Text\n\nLeitold D, Vathy-Fogarassy A, Abonyi J: Network distance-based simulated annealing and fuzzy clustering for sensor placement ensuring observability and minimal relative degree. Sensors (Basel). 2018; 18(9): pii: E3096. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBounova G: Octave networks toolbox. 2015. Reference Source\n\nGori F, Folino G, Jetten MS, et al.: MTR: taxonomic annotation of short metagenomic reads using clustering at multiple taxonomic ranks. Bioinformatics. 2011; 27(2): 196–203. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRuths J, Ruths D: Control profiles of complex networks. Science. 2014; 343(6177): 1373–1376. PubMed Abstract \| Publisher Full Text\n\nFreeman LC: A set of measures of centrality based on betweenness. Sociometry. 1977; 40(1): 35–41. Publisher Full Text" }	[ { "id": "48269", "date": "28 May 2019", "name": "Gilles Didier", "expertise": [ "Reviewer Expertise My main area of research is applied mathematics. I worked on biological networks (regulatory and interaction networks)." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis software tool article presents the Octave/MATLAB toolbox \"NOCAD\" for networks analysis. NOCAD implements various standard measures on networks and their nodes (density, betweenness, centrality...). Its main novelty stands in the ``dynamical-systems-based'' tools to analyse (static) networks that it provides. This article can be seen as a companion paper of the reference 191 from the same authors (It could be stated somewhere in the text). The ideas implemented in the toolbox are interesting, but I feel that their presentation has to be improved. First, I suggest you get your text proof-read by a native speaker. Second, many concepts used in the manuscript should be at least briefly recalled. Third, the article lacks a detailed description of the input (in the general sense, including user interaction) of the toolbox. The basic input seems to be the incidence matrix of a directed network but, if I well understood, the toolbox allows to set input and output nodes and has network design capabilities. Another major concern is that though one of the aims of the article is to draw attention of life science researchers, it does not present clearly how the toolbox can be used on biological networks and on which kind of biological networks it can be applied. Real biological examples should be actually more than welcome in the \"Use cases\" section, which contains only toy examples. Without such examples, I am afraid that the article fails to draw attention of its expected audience. In particular, I am wondering in what extent the toolbox can be used on standard biological datasets, e.g., PPI networks can have thousands vertices.\nPlease find below some detailed comments:\nAbstract The abstract has to be extended in order to give a better idea of the content of the article.\nIntroduction The introduction does not reflect well what follows and the topic of the paper. You should start by briefly introducing the general ideas of dynamical analysis of networks, precise in what sense you use the term \"driver node\", which is sometimes used in a biological context etc. First mention of dynamical system comes at the end of the second paragraph and could be developed (e.g. by briefly recalling ideas of [7]2) while biological examples may not need to be as much detailed. The description of previous software sounds good. It may take place a dedicated section.\nMethod The method section should recall the notions of controllability and observability and the general principles of ``dynamical analysis'' (it can be brief, something like the two first subsections of the section \"Results\" of [19]1). In the current version, I don't think that this part can be understood without reading your reference [19]1. You should state explicitly which are the inputs of your toolbox, the accepted formats etc. I found the paragraph about the modularity of the toolbox is not very useful. It could be shortened or even discarded. Subsection \"Operation\" (\"Installation\"?) is OK.\nUse cases If I well understood, the input is the incidence matrix A from which are computed the matrices B, C and D as illustrated with the first example. It should be interesting to apply the approach on a biological network, for instance a protein-protein interaction network or a regulatory network, in order to illustrate the interest of the results obtained in this context. In the second example (Figure 3), input and output nodes seem to be set arbitrarily in order to get a system complex-enough to show the function of the second module of the toolbox (incidence matrix is the same as that of the first example). Is it possible to illustrate the toolbox with an example in which the input and outputs are not fixed? I am actually concerned with the fact that the use cases do not really show how the toolbox can be used on a real dataset, which may not have natural input and outputs. Figures 4, 5 and 6 displays the results of the toolbox on the second example. These results are briefly discussed in the text which also recalls the definitions of some of the measures. Applying the toolbox on a real dataset would make the discussion of the results of the toolbox much easier to interpret. Moreover, a real dataset could provide evidence of the relevance of the results obtained by relying on the nature of nodes distinguished by the toolbox. The different types of nodes displayed in Figure 5 should be defined rather than just refer to [24]3. Could you please give an idea about what kind of biological networks can be studied by the means of your third module? This part is not very intuitive from a biological point of view.\n\nIs the rationale for developing the new software tool clearly explained? Partly\n\nIs the description of the software tool technically sound? Partly\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? No\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly", "responses": [ { "c_id": "4911", "date": "21 Oct 2019", "name": "Janos Abonyi", "role": "Author Response", "response": "Dear Reviewer, We are grateful for your useful remarks. In the following, we provide a detailed report about how we improved the paper based on your valuable comments and suggestions. We hope that the modifications have significantly improved the understandability of the paper. Sincerely yours,Janos Abonyi This software tool article presents the Octave/MATLAB toolbox \"NOCAD\" for networks analysis. NOCAD implements various standard measures on networks and their nodes (density, betweenness, centrality...). Its main novelty stands in the ``dynamical-systems-based'' tools to analyse (static) networks that it provides. This article can be seen as a companion paper of the reference 191from the same authors (It could be stated somewhere in the text). Thank you for your valuable remark. We have cited the works related to the functions of the toolbox. The sections entitled Introductionand Methodshave been rewritten and improved in order to highlight the applicability of the toolbox in the life sciences and provide additional contribution. The ideas implemented in the toolbox are interesting, but I feel that their presentation has to be improved. First, I suggest you get your text proof-read by a native speaker. Second, many concepts used in the manuscript should be at least briefly recalled. Third, the article lacks a detailed description of the input (in the general sense, including user interaction) of the toolbox. The basic input seems to be the incidence matrix of a directed network but, if I well understood, the toolbox allows to set input and output nodes and has network design capabilities. Thank you for this remark. We used Grammarly and asked a professional proofreader to correct the second version of the manuscript. We have tried our best to minimize the number of grammatical errors. At the beginning of the section entitled Methods, the representation of linear systems, definitions of structural controllability as well as observability, and introductions to the methodology in addition to maximum matching were included, therefore, a review of the cited literature was deemed unnecessary. The inputs of the modules were clarified in the section entitled Implementation. Another major concern is that though one of the aims of the article is to draw attention of life science researchers, it does not present clearly how the toolbox can be used on biological networks and on which kind of biological networks it can be applied. Real biological examples should be actually more than welcome in the \"Use cases\" section, which contains only toy examples. Without such examples, I am afraid that the article fails to draw attention of its expected audience. In particular, I am wondering in what extent the toolbox can be used on standard biological datasets, e.g., PPI networks can have thousands vertices. We are grateful for your valuable remark and have cited the latest articles to highlight the importance of the methodology in the field of the life sciences. In the section entitled Use cases, the example network was replaced by a frontal neural network of C. elegans from the standard biological dataset that has been comprehensively analysed and is sufficiently well-known to be regarded as an interesting example for scientists from the fields of the life as well as network sciences. Please find below some detailed comments:AbstractThe abstract has to be extended in order to give a better idea of the content of the article. Thank you for this remark. We have improved the abstract based on the reviews and extended it with the information provided to develop the introduction with regard to the content and the main goal of the paper, namely the applicability of the methodology to the life sciences. IntroductionThe introduction does not reflect well what follows and the topic of the paper. You should start by briefly introducing the general ideas of dynamical analysis of networks, precise in what sense you use the term \"driver node\", which is sometimes used in a biological context etc. First mention of dynamical system comes at the end of the second paragraph and could be developed (e.g. by briefly recalling ideas of [7]2) while biological examples may not need to be as much detailed. We are grateful for your useful remarks. We have overviewed the Introductionand rewritten it to draw attention to the importance of the methodology in the life sciences. Furthermore, in the section entitled Methods, we introduced the theoretical background to network-based structural controllability and observability analysis. The description of previous software sounds good. It may take place a dedicated section. Thank you for this remark, we decided to dedicate a separate section to the existing software. MethodThe method section should recall the notions of controllability and observability and the general principles of ``dynamical analysis'' (it can be brief, something like the two first subsections of the section \"Results\" of [19]1). In the current version, I don't think that this part can be understood without reading your reference [19]1. We are grateful for this useful suggestion. As previously mentioned, we have included a theoretical introduction to dynamical systems and their controllability as well as observability in the section entitled Methods. You should state explicitly which are the inputs of your toolbox, the accepted formats etc. We are grateful for your remark. The introduction to dynamical systems emphasises the input of the toolbox, namely the dynamical matrix of the state-space representation. In addition, in the section entitled Implementation, each of the inputs to the modules was described separately. I found the paragraph about the modularity of the toolbox is not very useful. It could be shortened or even discarded. We are grateful for your valuable suggestion and have deleted this paragraph. Subsection \"Operation\" (\"Installation\"?) is OK.Use casesIf I well understood, the input is the incidence matrix A from which are computed the matrices B, C and D as illustrated with the first example. It should be interesting to apply the approach on a biological network, for instance a protein-protein interaction network or a regulatory network, in order to illustrate the interest of the results obtained in this context. Thank you for your valuable remark. We have replaced the example network with the frontal neural network of C. elegans to draw attention to the easily understandable measures rather than the useful results. Therefore, the results from the analysis of the frontal neural network of C. elegans are presented, while in the manual the example network is still available. In the second example (Figure 3), input and output nodes seem to be set arbitrarily in order to get a system complex-enough to show the function of the second module of the toolbox (incidence matrix is the same as that of the first example). Is it possible to illustrate the toolbox with an example in which the input and outputs are not fixed? I am actually concerned with the fact that the use cases do not really show how the toolbox can be used on a real dataset, which may not have natural input and outputs. Since we replaced the example network with the frontal neural network of C. elegans, it is not necessary to apply the fixed inputs and outputs, therefore, we hope that this problem has been resolved. Figures 4, 5 and 6 displays the results of the toolbox on the second example. These results are briefly discussed in the text which also recalls the definitions of some of the measures. Applying the toolbox on a real dataset would make the discussion of the results of the toolbox much easier to interpret. Moreover, a real dataset could provide evidence of the relevance of the results obtained by relying on the nature of nodes distinguished by the toolbox. As the example network was changed, we hope this remark has been acted on. The different types of nodes displayed in Figure 5 should be defined rather than just refer to [24]3. Thank you for this invaluable remark. We have expanded this part of the paper and, with references, mentioned the importance of these types in biological networks. Could you please give an idea about what kind of biological networks can be studied by the means of your third module? This part is not very intuitive from a biological point of view. We are grateful for this useful suggestion. We have dedicated a paragraph to the importance of this module in the section entitled Implementation. Given the improved theoretical background, we hope that the presence of this module is more suitable. Is the rationale for developing the new software tool clearly explained?Partly As no widespread and approved tool for the network-based structural controllability and observability analysis of dynamical systems is known, we propose our toolbox for this purpose, and its applicability is introduced with the aid of a real network from a standard biological dataset. Is the description of the software tool technically sound?Partly We have improved the introduction to the theoretical background, therefore, the paper can be regarded as an independent work that introduces the advantages of the methodology and implemented toolbox. Are sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others?Partly We have stated the input used in this article and the toolbox, moreover, a well-known example was used in order of maximal reproducibility. Is sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool?No We have included a more detailed introduction to the theoretical background in order to provide sufficient information concerning the expected outputs and results. Are the conclusions about the tool and its performance adequately supported by the findings presented in the article?Partly We hope that the frontal neural network of C. elegans and its results represent well the applicability of the toolbox, and believe in the importance of such a novel toolbox, even though numerous publications utilise this methodology." } ] }, { "id": "48912", "date": "25 Jun 2019", "name": "Arthur Montanari", "expertise": [ "Reviewer Expertise Control theory", "control of networked", "nonlinear dynamics." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this paper the authors describe the NOCAD toolbox developed for the analysis of some aspects of networks as, for instance, structural controllability and observability. The topic of the paper is not only very interesting but also timely as issues relating to controllability and observability of networks could be fundamental in a number of practical situations. Hence to have a nice set of tools to analyze dynamical networks is welcome.\nThe toolbox described is useful in a specific range of problems. Here we focus on the controllability and observability properties of networks, as prompted by the title. The tools presented focus on structural controllability and observability of linear systems. The assumption of linearity is not mentioned in the abstract. As a matter of fact, the opening phrase of the last paragraph in the introduction should be copied to the abstract.\nThe nonlinear case and other definitions of controllability and observability, such as, dynamical and symbolical are not mentioned in the paper nor are handled by the toolbox. This is an important remark as it is now known that some algorithms underestimate the cardinality of the set of sensor nodes when applied to nonlinear systems.\nThe paper\nIn the second paragraph of the Introduction, the authors mention the importance of “determining the proper driver nodes”. We wonder if the average reader would know what that is. In clarifying this point the authors would like to address first what is “a” proper set of driving nodes (e.g. one that will guarantee full controllability). However, in order to specify “the” proper set, possibly some more detailed measure of controllability should be employed. For instance, suppose two sets of driving nodes S1 and S2, both with the same cardinality, such that the network is fully controllable either from S1 or from S2. Hence S1 is “a” proper set of driving nodes (assuming that by proper the authors are referring to full controllability) and S2 is another one. Now, it could well be that using S1 less energy is required to drive the network from one state x(ti) to another x(tf) when compared to S2. In that case, S1 and S2 are not totally equivalent. Structural controllability and observability are unable on their own to provide this distinction.\n\nFollowing the same vein, we wonder if the average reader of F1000Research would know the distinction of static and dynamic networks. A word about this would be profitable, especially because the toolbox refers mainly to the second class.\nIn some parts of the paper the authors refer to matrix A as the “state transition” matrix. In continuous-time this is incorrect. In discrete-time this is only correct for a transition time of one sampling period. Matrix A is called the “dynamical matrix”. State transition matrix is something else.\nRelevance to the journal\nThe MS introduction is written in such a way that it instigates interest from the target audience of this journal (mathematical biology). However, the techniques implemented in the toolboxes are usually found around the network science community, which includes several other fields that range from power systems to social networks. To show more coherence to this journal, the authors should provide some example of application on a “real-world network” under a biological context. This is also interesting to show how the results provided by the toolbox can be useful to draw conclusions under a practical context other than a “toy problem” as presented.\nBackground\nThe MS content is not sufficient to understand the techniques implemented in the toolbox. We know that the paper does not aim at providing extensive background, but some could be helpful. A reader coming from a control theory background might expect the toolbox to provide results based on Kalman’s definition of controllability and observability. However, the toolbox is based on Liu and coworker’s maximum matching algorithm which is based, in turn, on Lin’s structural definition (1974) (which is not even mentioned in the MS). Which definition of controllability and observability the toolbox is based on should be crystal clear in the main text and abstract. Moreover, the notions of structural controllability and observability should be presented to the reader and how they interplay with the more well-known notion in Kalman’s sense (e.g. Lin’s definition is only a necessary condition for Kalman’s definition). Although we think that the definitions of controllability and observability should be mentioned on the main text, the authors provide some background on the implemented maximum matching algorithms in the toolbox. This, however, should be mentioned explicitly in the MS.\nExamples\nThe paper furnishes examples to illustrate some of the features of the new toolbox. The network topology is the same in each example, which facilitates understanding and comparison. In some cases, more discussion would be welcome. For instance, in the last example (Figure 7) not all the indices are clearly defined to the user, for instance the critical nodes are given as x2, x4 and x7. From the context it seems that if any of these nodes is lost, the network would become uncontrollable, but this is not directly stated.\nIn referring to Figure 5 the authors speak in terms of Reachability matrices: Rc and Ro. It is clear from Figure 5 that Rc shows which nodes have a path to node i (the ith column). As for Ro, it seems that we should look row-wise instead of column-wise, is that right? Anyhow, we do not think it is adequate to say that Rc shows which nodes can be controlled from another one. The word “control” is perhaps too general. We would suggest just to say that Rc shows which nodes can be reached from another one.\nToolbox\nThe toolbox is subdivided in three modules. The first one, “network mapping module”, implements the maximum matching algorithm (and related modifications) to return a structurally controllable (observable) network with the smallest set of driver (sensor) nodes. This is a welcome feature, especially for a MATLAB environment, which we are unaware of any alternative.\nThe second module, “system characterization module”, provides several graph measures that are available on other MATLAB-based toolboxes, but are indeed useful to assess the network controllability and observability properties of a system. Thus, its presence is justifiable.\nThe third module, “improvements and robustness module”, is a set of functions which specifically implement previous results of the authors (e.g. Refs. 201, 212). It seems quite specific, but nevertheless the toolbox relevance is justifiable in great part for its module one.\nThe toolbox seems fast and no bugs were found in its implementations as far as the MATLAB environment is concerned. See further comments on some compatibility issues when using Octave.\nThe target audience of the toolbox\nThe toolbox is applicable to any kind of network (graph), be it directed or undirected, weighted or unweighted, and so on. However, although general, the authors should discuss in the MS what are the kinds of networks where a structural analysis of controllability and observability are more useful. For instance, a linearization of a power system model modelled by interconnected Kuramoto oscillators yield a dynamical matrix “A” whose corresponding adjacency graph is not only highly connected but also undirected. Consequently, the toolbox points out that only one driver node is needed to structurally control the network, independently of its size “n”(which is true in this case according to Lin’s definition). However, this does not give insight to the problem since basically any node can be chosen as a driver node, and only one node being sufficient seems quite unrealistic. The question is: For what kind of networks is the structural approach more interesting (and hence the toolbox)? It could be that the toolbox could be extra-helpful in the context of more sparse and directed networks (with higher hierarchy).\nSome nitpicking:\nSometimes it is not clear whether the authors refer to the dynamical matrix “A” or the adjacency matrix “A^T”. We recommend that different nomenclatures be used, such as “A” for the dynamical matrix and “A_{Adj}” for the adjacency matrix. This should be changed in the MS and manual.\nRegarding the installation of the NOCAD toolbox. We noticed that the NOCAD already comes with a “octave-network-toolbox” folder which does not have all the necessary functions to use Module 2. Thus, we had to download the “octave-network-toolbox-master” folder in Ref. 22 to have access to all needed functions. Is this necessary or is the NOCAD toolbox really missing some functions?\nSome statements in Section “Use Cases”, Paragraph 5, are redundant with the information already present in Fig. 4. Instead of repeating the same information, it might be more interesting to make some comments on how useful some of these network measures can be to design better and more robust networks from a control and observation point-of-view.\nThere are some issues in the Octave version of this toolbox. For instance, function “heatmaps” has some bugs in Octave but works well in MATLAB. This happens because the function “colormap” from Octave accepts the argument “hot” but not “Hot”. The authors should do a careful review of the Octave toolbox and check for further compatibility issues.\n\nIs the rationale for developing the new software tool clearly explained? Partly\n\nIs the description of the software tool technically sound? Partly\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly", "responses": [ { "c_id": "4912", "date": "21 Oct 2019", "name": "Janos Abonyi", "role": "Author Response", "response": "Dear Reviewers, We are grateful for your useful remarks. In the following, we provide a detailed report about how we improved the paper based on your valuable comments and suggestions. We hope that the modifications have significantly improved the understandability of the paper. Sincerely yours,Janos Abonyi In this paper the authors describe the NOCAD toolbox developed for the analysis of some aspects of networks as, for instance, structural controllability and observability. The topic of the paper is not only very interesting but also timely as issues relating to controllability and observability of networks could be fundamental in a number of practical situations. Hence to have a nice set of tools to analyze dynamical networks is welcome. The toolbox described is useful in a specific range of problems. Here we focus on the controllability and observability properties of networks, as prompted by the title. The tools presented focus on structural controllability and observability of linear systems. The assumption of linearity is not mentioned in the abstract. As a matter of fact, the opening phrase of the last paragraph in the introduction should be copied to the abstract. The nonlinear case and other definitions of controllability and observability, such as, dynamical and symbolical are not mentioned in the paper nor are handled by the toolbox. This is an important remark as it is now known that some algorithms underestimate the cardinality of the set of sensor nodes when applied to nonlinear systems. Thank you for this invaluable remark, we have emphasised in both the Abstractand Introductionthat the toolbox is applicable to the analysis of linear and linearized systems rather than nonlinear systems for which incorrect results may be suggested. The paper In the second paragraph of the Introduction, the authors mention the importance of “determining the proper driver nodes”. We wonder if the average reader would know what that is. In clarifying this point the authors would like to address first what is “a” proper set of driving nodes (e.g. one that will guarantee full controllability). However, in order to specify “the” proper set, possibly some more detailed measure of controllability should be employed. For instance, suppose two sets of driving nodes S1 and S2, both with the same cardinality, such that the network is fully controllable either from S1 or from S2. Hence S1 is “a” proper set of driving nodes (assuming that by proper the authors are referring to full controllability) and S2 is another one. Now, it could well be that using S1 less energy is required to drive the network from one state x(ti) to another x(tf) when compared to S2. In that case, S1 and S2 are not totally equivalent. Structural controllability and observability are unable on their own to provide this distinction. Thank you for your valuable remark, we have expanded the relevant part of the paper and emphasized that during optimisation the cardinality and energy demand should be minimised, while the configurations should provide structural controllability and observability. Following the same vein, we wonder if the average reader of F1000Research would know the distinction of static and dynamic networks. A word about this would be profitable, especially because the toolbox refers mainly to the second class. We are grateful for this suggestion, in the Introductionwe have clarified the nomenclature as well as system and network classes used in the manuscript. In some parts of the paper the authors refer to matrix A as the “state transition” matrix. In continuous-time this is incorrect. In discrete-time this is only correct for a transition time of one sampling period. Matrix A is called the “dynamical matrix”. State transition matrix is something else. Thank you for this invaluable remark, we have corrected the incorrect nomenclature. Relevance to the journal The MS introduction is written in such a way that it instigates interest from the target audience of this journal (mathematical biology). However, the techniques implemented in the toolboxes are usually found around the network science community, which includes several other fields that range from power systems to social networks. To show more coherence to this journal, the authors should provide some example of application on a “real-world network” under a biological context. This is also interesting to show how the results provided by the toolbox can be useful to draw conclusions under a practical context other than a “toy problem” as presented. We are grateful for this useful suggestion. We have improved the Introductionand, with the aid of references, introduced the networks that are already utilised in network-based structural controllability and observability analysis. In addition, the example problem has been replaced by the frontal neural network of C. elegans, a well-known standard biological dataset. Background The MS content is not sufficient to understand the techniques implemented in the toolbox. We know that the paper does not aim at providing extensive background, but some could be helpful. A reader coming from a control theory background might expect the toolbox to provide results based on Kalman’s definition of controllability and observability. However, the toolbox is based on Liu and coworker’s maximum matching algorithm which is based, in turn, on Lin’s structural definition (1974) (which is not even mentioned in the MS). Which definition of controllability and observability the toolbox is based on should be crystal clear in the main text and abstract. Moreover, the notions of structural controllability and observability should be presented to the reader and how they interplay with the more well-known notion in Kalman’s sense (e.g. Lin’s definition is only a necessary condition for Kalman’s definition).Although we think that the definitions of controllability and observability should be mentioned on the main text, the authors provide some background on the implemented maximum matching algorithms in the toolbox. This, however, should be mentioned explicitly in the MS. We are grateful for your useful remarks and have improved the section entitled Methods, moreover, extended it with the theoretical background to dynamical systems, definitions of structural controllability as well as observability, and the applied methods. The proposed works from the literature have been cited in the relevant sections and for this suggestion we would like to express our sincere gratitude. Examples The paper furnishes examples to illustrate some of the features of the new toolbox. The network topology is the same in each example, which facilitates understanding and comparison. In some cases, more discussion would be welcome. For instance, in the last example (Figure 7) not all the indices are clearly defined to the user, for instance the critical nodes are given as x2, x4 and x7. From the context it seems that if any of these nodes is lost, the network would become uncontrollable, but this is not directly stated. In referring to Figure 5 the authors speak in terms of Reachability matrices: Rc and Ro. It is clear from Figure 5 that Rc shows which nodes have a path to node i (the ith column). As for Ro, it seems that we should look row-wise instead of column-wise, is that right? Anyhow, we do not think it is adequate to say that Rc shows which nodes can be controlled from another one. The word “control” is perhaps too general. We would suggest just to say that Rc shows which nodes can be reached from another one. We are grateful for your valuable remark and have expressed the results provided by the toolbox as well as inserted a more thorough introduction to the results. According to the reachability matrices, control and reach derived from several dynamical systems overlap to some extent when the problem with regard to controllability was reduced to one concerning reachability. Toolbox The toolbox is subdivided in three modules. The first one, “network mapping module”, implements the maximum matching algorithm (and related modifications) to return a structurally controllable (observable) network with the smallest set of driver (sensor) nodes. This is a welcome feature, especially for a MATLAB environment, which we are unaware of any alternative. The second module, “system characterization module”, provides several graph measures that are available on other MATLAB-based toolboxes, but are indeed useful to assess the network controllability and observability properties of a system. Thus, its presence is justifiable. The third module, “improvements and robustness module”, is a set of functions which specifically implement previous results of the authors (e.g. Refs. 201, 212). It seems quite specific, but nevertheless the toolbox relevance is justifiable in great part for its module one. The toolbox seems fast and no bugs were found in its implementations as far as the MATLAB environment is concerned. See further comments on some compatibility issues when using Octave. We are grateful for your useful remarks and have highlighted in the section entitled Implementationthe importance of the third module for dynamical systems in general and biological systems specifically. The target audience of the toolbox The toolbox is applicable to any kind of network (graph), be it directed or undirected, weighted or unweighted, and so on. However, although general, the authors should discuss in the MS what are the kinds of networks where a structural analysis of controllability and observability are more useful. For instance, a linearization of a power system model modelled by interconnected Kuramoto oscillators yield a dynamical matrix “A” whose corresponding adjacency graph is not only highly connected but also undirected. Consequently, the toolbox points out that only one driver node is needed to structurally control the network, independently of its size “n”(which is true in this case according to Lin’s definition). However, this does not give insight to the problem since basically any node can be chosen as a driver node, and only one node being sufficient seems quite unrealistic. The question is: For what kind of networks is the structural approach more interesting (and hence the toolbox)? It could be that the toolbox could be extra-helpful in the context of more sparse and directed networks (with higher hierarchy). Thank you for your valuable remark. We believe no “better” types of networks for structural analysis are known since the importance of a node in a complex system from a structural point of view can be excessive or critical with regard to reliability, independent of the directedness in the network. The hardest aspect of these analyses is the evaluation of the results that should be provided by experts, moreover, this may cause undirected or directed and sparser or denser networks to be preferred. Some nitpicking: Sometimes it is not clear whether the authors refer to the dynamical matrix “A” or the adjacency matrix “A^T”. We recommend that different nomenclatures be used, such as “A” for the dynamical matrix and “A_{Adj}” for the adjacency matrix. This should be changed in the MS and manual. Thank you for this invaluable remark, we have used the nomenclature “A^T” to emphasise the transposition, i.e. the connection between the dynamical and adjacency matrices. Regarding the installation of the NOCAD toolbox. We noticed that the NOCAD already comes with a “octave-network-toolbox” folder which does not have all the necessary functions to use Module 2. Thus, we had to download the “octave-network-toolbox-master” folder in Ref. 22 to have access to all needed functions. Is this necessary or is the NOCAD toolbox really missing some functions? We are grateful for the thorough testing of the toolbox. Unfortunately, some function from the octave-networks-toolbox was excluded during its upload, but this has been corrected. Some statements in Section “Use Cases”, Paragraph 5, are redundant with the information already present in Fig. 4. Instead of repeating the same information, it might be more interesting to make some comments on how useful some of these network measures can be to design better and more robust networks from a control and observation point-of-view. Thank you for your remark, we have replaced the example network, therefore, the diagrams have been removed and the relevant parts improved. We hope that this problem has been resolved. There are some issues in the Octave version of this toolbox. For instance, function “heatmaps” has some bugs in Octave but works well in MATLAB. This happens because the function “colormap” from Octave accepts the argument “hot” but not “Hot”. The authors should do a careful review of the Octave toolbox and check for further compatibility issues. Thank you for your remark. Interestingly, our system did not alert us to this problem. We have corrected the source code as suggested since both parameters were applied to our configurations, however, we kindly ask you to check your version of Octave as the problem can also be caused by the use of an older version of this program." } ] } ]	1	https://f1000research.com/articles/8-646
https://f1000research.com/articles/7-1891/v1	04 Dec 18	{ "type": "Research Article", "title": "Factors associated with the utilization of long-acting reversible contraceptives among family planning clients at the Pameungpeuk Rural Hospital, Indonesia", "authors": [ "Achmad Kemal Harzif", "Ana Mariana", "Devi Marischa Malik", "Melisa Silvia", "Bara Tracy Lovita", "Ana Mariana", "Devi Marischa Malik", "Melisa Silvia", "Bara Tracy Lovita" ], "abstract": "Background: Uncontrolled population development can prompt an assortment of populace issues and can be one of the reasons for increasing maternal death rates. The utilization of contraceptives in Indonesia was progressively dominated by injectable contraceptives and pill contraceptives in 2015 (52.21% and 24.36%, respectively). However, the rate of termination of the use of short-acting contraceptives by family planning clients was higher than other methods, therefore the use of short-acting contraceptive is not effective enough for use. In anticipating the decreased use of short-acting contraceptives while also seeking to control population growth, the National Family Planning Program in Indonesia is encouraging the use of long-acting reversible contraceptives (LARCs). Pameungpeuk is a region which has the second largest population, with the highest total fertility rate in South-West Java. The proportion of active users of LARCs in Pameungpeuk is also very low (10.66%). This study therefore aimed to analyze factors associated with the utilization of LARCs among family planning clients at the Pameungpeuk Rural Hospital. Methods: This study used a cross-sectional design with systematic random sampling. The sample group in this study was 84 family planning clients. We performed statistical analyses using a chi-square test. Results: We found significant associations between the age of women (p=0.024), the cost of contraception (p=0.022), knowledge (p=0.042), beliefs (p=0.002), skill of health workers (p=0.008) and support from health workers (p=0.014). However, education (p=0.212), family income (p=0.087), attitude (p=0.593), exposure to information on LARCs (p=0.378), support from partners (p=0.094), support from friends (p=0.414) and the support of community leaders (p=0.367) had no significant association with the utilization of LARCs. Conclusions: These findings highlight a critical need for improved education among family planning clients at the Pameungpeuk rural hospital regarding the use of LARCs for both medical and elective reasons.", "keywords": [ "Family Planning", "Long-acting reversible contraceptives (LARCs)", "Pameungpeuk Rural Hospital" ], "content": "Introduction\n\nAccording to the World Population Data Sheet (2013), Indonesia is the fourth most populous country in the world, with a population estimate of 249 million. Among Association of South East Asian Nations (ASEAN) countries, Indonesia has both the largest area and largest population. In terms of birth rate, Indonesia is far above the 9 other member countries with a total fertility rate (TFR) of 2.6, while the average TFR of ASEAN countries is 2.41.\n\nIncreased and uncontrolled population growth can be one of the reasons behind increasing maternal death rates. Demographic and Health Survey (DHS) information from 2007 set up the Indonesian maternal death rates as 228 deaths for every 100,000 live births2. It is estimated that the global maternal death rates will be 195 deaths per 100,000 live births, while the proportion of maternal death in females reproductive age (PMDF) estimated will be 264–285 deaths per 100,000 live births in 20153. The maternal death rate in Indonesia is high, and was recently estimated as 126 deaths per 100,000 live births by the World Bank. As an effort to improve maternal health and also achieve the fifth Millennium Development Goal, it required a reduction in maternal death by 9.5% every year from 2005 to 2015, but in reality this goal was not achieved between 2005–20083,4.\n\nLow contraceptive use is associated with high maternal morbidity and mortality rates, unwanted pregnancies, short birth interval and higher risk of obstetric and newborn complications5. One of the Indonesian government's efforts to reduce population growth and its associated problems, including high mortality rates, are family planning programs promoting contraceptive use, which educate patients and clients regarding family planning and contraception itself4. Based on data from the WHO, compared to other ASEAN countries, the use of contraceptives in Indonesia (61% of the population) already exceeds the ASEAN average (58.1%). However, it is still lower than that in Vietnam (78%), Cambodia (79%) and Thailand (80%)6. Based on data from the DHS (2012), the use of contraception in Indonesia is dominated by injectable contraceptives (32%) and the pill (14%).\n\nIn 2010–2014, the National Family Planning Program in Indonesia directed more efforts towards promoting the use of long-acting reversible contraceptives (LARCs). LARCs are methods of contraception that are known to be highly effective, as they can provide protection from the risk of pregnancy for a period of up to 10 years, depending on the type used6,7.\n\nThe advantages of using long-term contraceptive methods are their convenience, comparatively low cost over time and predominant effectiveness in preventing pregnancy compared to short-acting methods. LARCs include implants, intra uterine devices (IUDs), male operation methods (MOPs) and female operation methods (MOWs)8. The use of LARCs is proposed as an important strategy to reverse undesirable maternal health consequences in developing countries9,10. Use of injectable has been predominantly high while IUD utilize has rapidly low in developing countries11,12. Based on data from the DHS (2012), the proportion of LARC users in Indonesia declined between 1994 and 2012. In 2012, the total number of LARC users was 10.6%, while the national target was 27.5%6.\n\nBased on the 2015 results of the State Ministry for Population and National Family Planning Coordinating Board (BKKBN), contraceptive use in Indonesia is increasingly dominated by short-acting contraceptives such as injectables (52.21%) and the pill (24.36%). The total number of LARC users in 2015 was (18.17%), while the majority used short-acting contraceptives (81.83%)13. Pameungpeuk is a region with the second largest population, and has the highest TFR in South-West Java. The proportion of active users of LARCs in Pameungpeuk is also considerably low (10.66%)14,15.\n\nThere are many factors that influence the lack of availability and access to LARCs method. These factors include fertility, lack of knowledge and other reasons related to methods that can act as barriers to the use of LARCs16. The factors than can inhibit access to the use of LARCs are higher costs for individuals, lack of trained health providers, short acting methods are widely available in rural areas, long distance to the clinic or hospital and medical barriers17,18. Widespread of myths and misunderstandings about LARCs and the lack of knowledge and skills of health workers can also influence access to the use of LARCs methods19,20.\n\nFactors that are suspected to be related to the utilization of LARCs are age, education, family income, cost of contraception, knowledge, beliefs, attitude of the family planning client, exposure to information on LARCs, contraceptive related skills, support from partners, friends, health workers, and also support from leaders in the community. There are, to our knowledge, no studies that have documented the factors associated with the lack of use of LARCs in the Pameungpeuk region. This study therefore aims to assess the factors associated with the utilization of LARC methods among family planning clients in the Pameungpeuk Rural Hospital, West Java, Indonesia.\n\n\nMethods\n\nThis was a hospital-based cross-sectional study. Eligible study participants were women seeking family planning services during the study period from the 15th of October, 2015 to the 15th of February 2016 at the Pameungpeuk Rural Hospital, which is a health institution that serves over 6000 annual pregnancies in the south of Garut, West Java Province, Indonesia.\n\nAn interviewer-administered questionnaire (Supplementary File 1) was used to collect data for this study. The questionnaire sought information on demographic characteristics, socio-economic status, reproductive history and utilization of modern contraception. Questionnaire items were developed to assess factors associated with the utilization of LARC methods, and included knowledge, beliefs, attitude, exposure to information on LARCs, skills of the health workers, and support from partners, friends, health workers and community leaders. We also sought to foster awareness among midwives and nurses who administer family planning services at the health institution and were asked to administer questionnaires; they were sensitized to the purpose and objectives of the study, and how to administer the questionnaire.\n\nThe interviewer-administered questionnaire (Supplementary File 1) was conducted with consenting consecutive clients as they accessed family planning services until the desired sample size was achieved. The study used a simple random sampling technique to avoid bias. The inclusion and exclusion criteria were as follows: women between 15–49 years who attended the family planning clinic at the Pameungpeuk Rural Hospital from the 15th of October, 2015 to the 15th of February, 2016. Those who did not give consent, those who could not respond to the interview questionnaire or those who incompletely filled out the questionnaire were excluded. The sample size for the study was calculated using the prevalence of contraceptive use in Pameungpeuk (12% for the rural region) from the Indonesian DHS 2013 report. The formula below was used:\n\n\n\nWhere n is the minimum sample size, Z is the standard normal deviation, set at 1.96, p is the prevalence of contraceptive use in the region (0.12), q is 1-p (0.88). The degree of accuracy was set at 0.05. The sample size was calculated to be 84 women of reproductive age.\n\nAfter all the necessary data was collected, the data was coded on pre-arranged coding sheets by the principal investigator. Statistical analyses and data entry were performed using the Statistical Package for the Social Sciences (SPSS), version 20.0. We performed statistical analyses using a chi-square test. Descriptive analysis was used to describe the study sample and the results are presented in tables.\n\nThis study was approved by The Ethics Committee of Faculty of Medicine, University of Indonesia under number 245/UN2.F1/ETIK/2015. The research proposal was submitted and reviewed by the Research Committee, whereby permission was granted to conduct the research. A written letter of consent was submitted to the health institution to seek permission to conduct this study, which was also granted. Privacy and confidentiality of the clients' information was observed through the use of data collection with coded identification numbers. Written informed consent was obtained from all participants prior to participation.\n\n\nResults\n\nA total of 84 family planning clients at the Pameungpeuk rural hospital participated in this study. As shown in Table 1, the majority of participants were more than 30 years old (61.90%). More than half of participants had their first marriage and first childbirth between the ages of 20 and 30 years (55.95% and 63.10%, respectively). Most of the participants had a low level of education, i.e. not beyond graduation from elementary school (32.14%). A majority of participants were Muslim (96.42%). In addition, more than half of participants (69.05%) had a low family income (i.e. a monthly income of less than 200 USD), and they thought that cost of contraceptives in Pameungpeuk was quite expensive (53.58%). Utilization of modern contraceptives at the Pameungpeuk rural hospital was dominated by the use of short-acting contraceptives (Non-LARCs), namely injectable contraceptives (59.52%) and pills (15.47%), while the use of LARCs was just less than 30% [IUD (16.66%) and implant (8.33%)]. (Table 1)\n\nLARCs, long-acting reversible contraceptives; LAM, lactational amenorrhea method.\n\nFactors that are suspected to be related to the utilization of LARCs were age, education, family income, cost of contraception, knowledge, beliefs, attitude of the family planning client, exposure to information about LARCs, contraceptive-related skills and support from partners, friends, health workers and leaders in the community. As shown in Table 2, the results of a bivariate analysis of factors revealed that age, cost of contraception, knowledge, beliefs, skills of health workers, and support from health workers were variables that were significantly associated with the utilization of LARCs. Based on the results of chi-square statistical analyses, of the 13 variables suspected as factors associated with utilization of LARCs, only 6 variables were significantly associated with the utilization of LARCs, and were as follows: age (p=0.024), cost of contraception (p=0.022), knowledge (p=0.042), beliefs (p=0.002), skills of health workers (p=0.008) and support from health workers (p=0.014). The other 7 variables, namely education, family income, attitude, exposure to information about LARCs, support of partners, friends and community leaders, were not significant associated with the utilization of LARCs at the Pameungpeuk Rural Hospital (Table 2).\n\n\nDiscussion\n\nThis study reveals a relationship between maternal age and LARC utilization. At <20 years or >30 years, family planning participants generally prefer high-effectiveness contraceptives such as the IUD, pill, or injections. These findings are consistent with previous research conducted by Tunnisa et al. and by Soppeng et al. in Pangkep, who found a relationship between maternal age and type of contraceptive use21,22. The results of this study are also consistent with those of Ama et al., who conducted a study in Botswana and found that the age of respondents is significantly associated with contraceptive use23.\n\nThe results of our data analysis show that there is no relationship between educational level and LARC use among family planning clients. This suggests that a high level of education does not affect the participants’ decision of which type of contraception to use; respondents with both low and high levels of education are aware of the importance and benefits of contraception, and have been likely made aware by health workers or other sources. The results of this study are also consistent with research conducted by Wa Ode in Pasarwajo, Buton, Indonesia, which found no relationship between the level of education and contraceptive use24.\n\nThe income of a family is closely related to their attitude towards contraceptive use; a family’s income is one of the factors that influences their acceptance towards, and the decision to use, new contraceptive innovations. Our results showed that respondents who used LARC and non-LARC contraceptive methods were mainly low-income. Among non-LARC users, only a small percentage had high incomes. This indicates that the desire of couples to accept family planning is still high, despite their low income, as in terms of cost, non-LARC contraceptives tend to be cheaper than LARCs. In this study, there was no significant relationship between family income and LARC contraceptive use (p=0.087). It can therefore be assumed that the level of family income does not affect the purchasing power of respondents to buy contraceptives. This research is inconsistent with research conducted by Maiharti et al., who found a significant relationship between family income and the use of contraceptive methods25.\n\nOur results show that there was a significant relationship between the cost of contraception and the use of hormonal contraception. Thus it can be said that the costs incurred for contraceptives are related to the choice of which contraception to use; the cost of non-LARCs tends to be less than that of LARCs. This study was consistent with research conducted by Aryanti et al., which showed a relationship between contraceptive costs and contraceptive selection26. The results of this study are also in accordance with research conducted by Fienalia et al. at the Pancoranmas Health Center, Depok, which found a relationship between the affordability of contraceptive costs and the use of LARCs27. Hartanto et al. found that there was a correlation between the use of contraceptive with support from husband or their partner28. Under ideal circumstances, couples should jointly choose the best method of contraception, mutually cooperate in their use, jointly pay for their expenses and take into account any dangerous side-effects. The results of statistical analyses showed no relationship between a partner’s support and the use of LARCs. Therefore, it can be assumed that a partners’ support does not affect the choice of the woman to seek family planning services and to continuously use contraceptives. This study was inconsistent with research conducted by Rizali et al., which found a relationship between support from the partner and the use of pill-based contraceptives29. This research is also inconsistent with a study conducted by Setyawati in the Mamojang district of Makassar City, Indonesia, which found that support from the partner will encourage participation in family planning30.\n\nThe results of our statistical analysis indicate that there is a relationship between the information provided by health-care personnel and the use of LARCs. This means that family planning clients already have information regarding contraceptive use from other sources. Others sources are from doctors, books, schools, television or friends. Our results showed that family planning LARC users were more likely to have not received information from family planning officers, while non-LARC users were more likely to have received information from family planning officers. This shows the lack of information dissemination from family planning officers to the community, or the lack of active support from health workers at the Pameungpeuk hospital towards the use of LARC methods, with the end result being a lack of information among the public. Based on information from one midwife who served in the family planning service, it is suspected that the health officer in charge is less active in conducting counseling regarding family planning by LARC methods. This research was consistent with Siddik et al., which found that there is a relationship between information dissemination and support from the family planning officers with the choice of contraceptive method. This is also consistent with research conducted by Rizali in the Mattoangin district of Makassar City, Indonesia, which showed a relationship between information dissemination by family planning officers and contraceptive use29,31.\n\n\nConclusions\n\nThis study found a significant relationship between LARC use and the age of women (p=0.024), cost of contraception (p=0.022), knowledge (p=0.042), beliefs (p=0.002), skills of health workers (p=0.008) and support from health workers (p=0.014). Education (p=0.212), family income (p=0.087), attitude (p=0.593), exposure to information about LARCs (p=0.378), support from partner (p=0.094), support from friends (p=0.414) and support from community leaders (p=0.367) all had no significant relationship with LARC utilization. Therefore, improving the skills of health workers related to contraception needs to be developed, besides that family planning acceptors are expected to play an active role in any activities related to contraception as well, especially in the use of long-term contraceptive methods at Pameungpeuk rural hospitals. Related sectors should seek to increase community knowledge regarding contraception and clarify of myths regarding the use of long-term contraceptive methods.\n\nThese findings highlight a critical need to improve partners’ understanding of the importance of using LARC methods, to support females in the use of contraception. This also requires increased knowledge dissemination by health-care workers. We recommended that family planning officers improve counseling to mothers and encourage them to remain active contraception users. They must also aid in the community’s understanding of the importance of contraceptive use, especially the utilization of long-term contraceptive methods.\n\n\nData availability\n\nDataset 1. All raw data and demographic information obtained from subject during the present study. DOI: https://doi.org/10.5256/f1000research.15755.d22802632.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting of this work.\n\n\nAcknowledgements\n\nThis work was performed at the Department of Obstetrics and Gynecology, Pameungpeuk Public Hospital, Garut. The authors wish to thank all staff at the Pameungpeuk Public Hospital Garut, especially for family planning provider sersvices, for supporting this study. We also would like to thank to Adele Tufford through AuthorAid for helping us improve our manuscript by copyediting and giving us thoughtful comments.\n\n\nSupplementary material\n\nSupplementary File 1. Questionnaire and informed consent form used in the present study.\n\nThe original Indonesian document is included alongside and English translation.\n\nClick here to access the data\n\n\nReferences\n\nprb.org [Internet]. Washington (DC): Population Reference Bureau; [cited 2018 May 11]. Reference Source\n\ndhsprogram.com [Internet]. USA: USAID, The DHS Program (Demographic and Health Surveys); [cited 2018 May 11]. Reference Source\n\nBappenas: Rancang Bangun: Percepatan Penurunan Angka Kematian Ibu Untuk Mencapai Sasaran Millennium Development Goals. Jakarta: Kementerian Perencanaan Pembangunan Nasional/Badan Perencanaan Pembangunan Nasional (Bappenas). 2007. Reference Source\n\nMinistry of Health, Indonesia: Strategies to Accelerate the Reduction of MMR. Jakarta: Ministry of Health. 1998.\n\nNalwadda G, Mirembe F, Byamugisha J, et al.: Persistent high fertility in Uganda: young people recount obstacles and enabling factors to use of contraceptives. BMC Public Health. 2010; 10: 530. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nState Ministry for Population and National Family Planning Coordinating Board (BKKBN): Basic Information: Family Planning Movement, Prosperous Family Development. Jakarta, Indonesia: BKKBN, 2012.\n\nState Ministry for Population and National Family Planning Coordinating Board (BKKBN): Basic Information: Family Planning Movement, Prosperous Family Development. Jakarta, Indonesia: BKKBN, 2010.\n\nWorld Health Organization: From Evidence to Policy: Expanding Access to Family Planning: Strategies to increase use of long-acting and permanent contraception. Geneva, Switzerland: World Health Organization; 2012. Reference Source\n\nUN: Contraceptive Commodities for Women’s Health, in Key Data and Findings. New York: United Nations Commission on Life-Saving Commodities for Women and Children, 2012; 1–29. Reference Source\n\nAnguzu R, Tweheyo R, Sekandi JN, et al.: Knowledge and attitudes towards use of long acting reversible contraceptives among women of reproductive age in Lubaga division, Kampala district, Uganda. BMC Res Notes. 2014; 7(1): 153. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDarroch JE, Singh S: Trends in contraceptive need and use in developing countries in 2003, 2008, and 2012: an analysis of national surveys. Lancet. 2013; 381(9879): 1756–1762. PubMed Abstract \| Publisher Full Text\n\nBertrand JT, Sullivan TM, Knowles EA, et al.: Contraceptive method skew and shifts in method mix in low- and middle-income countries. Int Perspect Sex Reprod Health. 2014; 40(3): 144–153. PubMed Abstract \| Publisher Full Text\n\nWijayanti N, Thaweesit S, Sunpuwan M: Contraceptive use among married adolescent women in Indonesia. J Health Res. 2015; 29(5): 323–31. Publisher Full Text\n\nMinistry of Internal Affairs: Rencana Pembangunan Jangka Menengah Daerah Kabupaten Garut tahun 2014-2019 (Medium-Term Development Plan for Garut District in 2014-2019). Garut: District Government Regulation of Garut No.3; 2014; 2014. Reference Source\n\nBiro Pusat Statistik Kabupaten Garut (Central Bureau of Statistic Garut): Survey Potensial Desa (PODES), Biannual Report. BPS, Garut. 2016.\n\nMbizvo MT, Phillips SJ: Family planning: choices and challenges for developing countries. Best Pract Res Clin Obstet Gynaecol. 2014; 28(6): 931–43. PubMed Abstract \| Publisher Full Text\n\nAlemayehu M, Belachew T, Tilahun T: Factors associated with utilization of long acting and permanent contraceptive methods among married women of reproductive age in Mekelle town, Tigray region, north Ethiopia. BMC Pregnancy Childbirth. 2012; 12(1): 6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBulto GA, Zewdie TA, Beyen TK: Demand for long acting and permanent contraceptive methods and associated factors among married women of reproductive age group in Debre Markos Town, North West Ethiopia. BMC Womens Health. 2014; 14(1): 46. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRavenholt AS, Sulzbach S, Banke K: Assessing The Commercial Viability of Long-Acting And Permanent Contraceptive Methods. 2009. Reference Source\n\nOsei I, Birungi H, Addico G, et al.: What happened to the IUD in Ghana? Afr J Reprod Health. 2005; 9(2): 76–91. PubMed Abstract \| Publisher Full Text\n\nTunnisa R: Faktor-Faktor yang Berhubungan dengan Pemilihan Alat Kontrasepsi di Wilayah Kerja Puskesmas Takalala Kecamatan Marioriwawo Kecamatan Soppeng Tahun 2010 (Skripsi). Fakultas Kesehatan Masyarakat Universitas Hasanuddin, Makassar. 2010.\n\nZainuddin E: Faktor yang Berhubungan dengan Pemilihan Metode Kontrasepsi Efektif Terpilih (MKET) Pada Akseptor KB di Kelurahan Tonasa Kecamatan Balocci Kabupaten Pangkep Tahun 2012 (Skripsi). Fakultas Kesehatan Masyarakat Universitas Hasanuddin, Makassar. 2012. Reference Source\n\nAma NO, Oucho JO: A Multivariate Approach to Determinant of Contraceptive Use Among Migrants and Refugees in Bostwana. Jurnal of Family Welfare. 2007; 53(2): 26–42. Reference Source\n\nWa Ode, et al.: Faktor yang Berhubungan dengan Penggunaan Metode Kontrasepsi Hormonal Pada Akseptor KB di Kelurahan Pasarwajo Kecamatan Pasarwajo Kabupaten Buton Sulawesi Tenggara. (Skripsi). 2013. Reference Source\n\nMaiharti: Hubungan Tingkat Pengetahuan, Pendidikan dan Pendapatan dengan Penggunaan Metode Kontrasepsi Pada PUS di Kecamatan Jenu dan Kecamatan Jatirogo Kabupaten Tuban (Skripsi). Universitas Negeri Surabaya, Surabaya. 2012.\n\nAryanti: Faktor-Faktor yang Berhubungan dengan Pemilihan Kontrasepsi Tubektomi di Kelurahan Jongaya Kecamatan Tamalate Kota Makassar Tahun 2010 (Skripsi). Fakultas Kesehatan Masyarakat Universitas Hasanuddin, Makassar. 2010.\n\nFienalia RA: Faktor-Faktor yang Berhubungan dengan Pemilihan Metode Kontrasepsi Jangka Panjang (MKJP) di Wilayah Kerja Puskesmas Pancoran Mas Kota Depok Tahun 2012 (Skripsi). Fakultas Kesehatan Masyarakat Universiatas, Indonesia, Jakarta. 2011.\n\nHartanto H: Keluarga Berencana dan Kontrasepsi. Jakarta: Pustaka Sinar Harapan. 2004. Reference Source\n\nRizali MI, Ikhsan M, Salmah AU: Faktor yang Berhubungan Dengan Pemilihan Metode Kontrasepsi Suntik di Kelurahan Mattoangin Kecamatan Mariso Kota Makassar. Media Kesehatan Masyarakat Indonesia. 2016; 9(3): 176–83. Reference Source\n\nSetiawaty L: Faktor-Faktor yang Berhubungan dengan Pemilihan Jenis Kontrasepsi Pil di Kecamatan Mamojang Kota Makassar Tahun 2006 (Skripsi). Fakultas Kesehatan Masyarakat Universitas Hasanuddin, Makassar. 2007.\n\nSiddik, Fajrita D, Hasnawati: Faktor yang Berhubungan dengan Pengambilan Keputusan Wanita Usia Subur (WUS) untuk BerKB IUD di Poli Kebidanan RSAL DR. MINTOHARDJO (Skripsi). Universitas Pembangunan Nasional Veteran, Jakarta. 2009.\n\nHarzif AK, Mariana A, Malik DM, et al.: Dataset 1 in: Factors associated with the utilization of long-acting reversible contraceptives among family planning clients at the Pameungpeuk Rural Hospital, Indonesia. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15755.d228026" }	[ { "id": "41452", "date": "19 Dec 2018", "name": "Luis Bahamondes", "expertise": [ "Reviewer Expertise Contraception" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn my opinion, the manuscript submitted by Harzif and co-workers entitled: Factors associated with the utilization of long-acting reversible contraceptives among family planning clients at the Pameungpeuk Rural Hospital, Indonesia” is important for readers for several reasons.\nIndonesia is one of the countries with a large population, high total fertility rate (TFR) and high maternal morbidity and mortality. Furthermore, although the use of modern contraceptives is increased in the last decades, it is mainly based on the use of short acting reversible contraceptives (SHARC), mainly injectables and pills (almost 75%). It is well established that the use of SHARC methods are associated to early discontinuation and high contraceptive failure when compared to users of long-acting reversible contraceptives (LARCs; intrauterine devices and implants). For this reason the government is encouraging the use of LARC methods. The authors showed data from Pameungpeuk, an Indonesian region which has the second largest population, with the highest TFR in South-West Java and with low use of LARCs (almost 10%).\nThe main findings of the study were that the use of LARCs were significantly associated with age of women, the cost of contraception, knowledge, beliefs, skill of health workers, and support from health workers.\nThe main limitation of the study was that it was a cross-section design and the number of interviewed women, which cannot allow to generalized the results to a large country like Indonesia. Additionally, it could encourage others to assess other regions to evaluate also the use of LARCs.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "46943", "date": "23 May 2019", "name": "Douglas Storey", "expertise": [ "Reviewer Expertise Reproductive health & family planning", "health communication research", "program evaluation", "over 30 years of work on Indonesian family planning program design and evaluation." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript provides a useful, localized analysis of contraceptive method choice in a particular area of West Java, but fails to provide adequate details about how key variables were operationalized, making it difficult to judge the quality of the analysis and some of the conclusions reached. The manuscript could be indexed with major revisions.\n\nDetailed comments and questions: (p1) In the Abstract: “...the use of short-acting contraceptive is not effective enough for use.”\nContraceptive should be plural (contraceptives). “Not effective enough for use” is an overstatement. Injectables and oral pills are highly effective at preventing pregnancy when used properly. They are, however, less efficient than longer acting or permanent methods when used for longer term spacing or limiting of childbirth because it is easy to skip a treatment or take a short break from use for economic or other reasons, which can result in unintended pregnancy. The authors should clarify the distinction between effectiveness of the methods and efficiency or practicality with regard to achieving fertility objectives.\n(p3, column 2, 1st full paragraph): “Based on the 2015 results...”\nAuthors should consider updating these contraceptive use statistics with the more recent results from the 2017 Indonesia Demographic & Health Survey (Survei Demografi dan Kesehatan Indonesia 2017).\n(p3, column 2, 2nd full paragraph): “Widespread of myths and misunderstandings...”.\nDelete word “of” “Myths” is a poorly defined word that is commonly misused to refer to negative beliefs or attitudes about contraceptive methods. Saying that someone believes in a “myth” suggests that they are superstitious and ignorant. It is more appropriate and more accurate to simply say that some people have “negative beliefs and misunderstandings” about LARCS.\n\n(p3, column 2, Section on Data Collection):\nMy most serious concerns about this manuscript regard the lack of clarity about how some key variables are conceptualized and measured. Authors should provide full descriptions of how variables were operationalized in the interview and/or manipulated (coded or recoded) for analysis. Specifically:\nNowhere in the manuscript is there a description of how “knowledge” was conceptualized and measured. Knowledge of what? Of methods available? Of how they work? Of contraindications or side effects? Of how long they can be used? Without this description it is impossible for the reader to know what “Good” or “Bad” knowledge means and therefore to be able to understand how or why his variable might be related to use of LARCs. Similarly, nowhere do the authors describe how “beliefs” and “attitudes” were conceptualized and measured. What beliefs and attitudes were asked about in the interview? Was there a single measure of belief or were multiple beliefs asked about and used to construct a knowledge scale or index? Were respondents asked about the strength of their belief in certain things about FP methods (e.g., “How strongly do you agree with each of the following statements?”) or simply asked, “Do you believe X, yes or no?). What attitudes toward contraceptive methods were assessed in the interview? How were they used to determine if attitudes were positive or negative? The manuscript needs much more detail about how these important independent variables were measured. Similarly, how was “Skills of health workers” conceptualized and measured? What skills were considered? Counseling skills? Clinical skills? How were some workers categorized as Skilled and others as Unskilled? It is impossible to know what this variable means in Table 2.\n\n(p5, 1st column):\nAuthors find that (perceived) cost of contraception has a significant negative effect on contraceptive use. It may be worthwhile to note that this study was (I believe) done before the new National Health Insurance scheme (Jaminan Kesehatan Nasional/Badan Penyelenggara Jaminan Sosial) was rolled out and widely available. National health insurance has, in recent years, overcome many of the cost barriers to contraceptive use, including use of LARCS.\n(p5, Table 2):\nI would find it very helpful to see cell percentages in the Utilization of Contraceptive columns of the table.\n(p7, 1st column, 2nd paragraph): “The results of our statistical analysis indicate that there is a relationship between the information provided by the health-care personnel and the use of LARCs.”\nThis comment is related to my concern above about how “health worker skills” was measured. How did the interview determine what information health workers provided to their clients? If the client reported that the health worker did not provide information about LARCS, is that what results in categorization of the health worker as Unskilled? This needs clarification.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4667", "date": "24 May 2019", "name": "Douglas Storey", "role": "Reviewer Response", "response": "Dear Authors - When I reviewed your manuscript, I overlooked the fact that the questionnaire was available through a link on the webpage. I can see now how the knowledge, attitudes, beliefs, health worker skills and other independent variables were presented in the questionnaire, but I still don’t know how the multiple measures for knowledge (Q12-Q20) were combined to create “good” vs “bad” knowledge. Same concern for beliefs (Q21-Q26), attitudes (Q17-Q36), exposure (Q37, multiple response) and health worker skills (Q38-Q41). Other independent variables were also constructed from multiple items in the questionnaire. In each case, I recommend that you explain how the analytic variables were created from multiple questionnaire items." }, { "c_id": "4852", "date": "17 Sep 2019", "name": "Achmad Kemal", "role": "Author Response", "response": "Dear Douglas Storey, Thank you for the review. We uploaded the revision of our manuscript. The text has been updated to address all of the points. The abstract is updated to clarify the efficiency of short-acting contraceptive. The prevalence of contraceptive methods section has been updated to describe more recent result from the State Ministry for Population and National Family Planning Coordinating Board (BKKBN) in 2017. Data collection section has been update to provide adequate details about how some key variables were conceptualized, measured and operationalized in the interview and/or manipulated (coded or recoded) for analysis. In Table 2, the percentages in the utilization of contraceptive has been added in the columns of table. We carefully have modified the manuscript according to the your instructions. Thank you for your considerations, we looking forward to any helpful feedback. Best Regards,Achmad Kemal" } ] } ]	1	https://f1000research.com/articles/7-1891
https://f1000research.com/articles/8-358/v1	01 Apr 19	{ "type": "Research Article", "title": "Adaptation, fitness landscape learning and fast evolution", "authors": [ "John Reinitz", "Sergey Vakulenko", "Dmitri Grigoriev", "Andreas Weber", "Dmitri Grigoriev", "Andreas Weber" ], "abstract": "We consider evolution of a large population, where fitness of each organism is defined by many phenotypical traits. These traits result from expression of many genes. Under some assumptions on fitness we prove that such model organisms are capable, to some extent, to recognize the fitness landscape. That fitness landscape learning sharply reduces the number of mutations needed for adaptation. Moreover, this learning increases phenotype robustness with respect to mutations, i.e., canalizes the phenotype. We show that learning and canalization work only when evolution is gradual. Organisms can be adapted to many constraints associated with a hard environment, if that environment becomes harder step by step. Our results explain why evolution can involve genetic changes of a relatively large effect and why the total number of changes are surprisingly small.", "keywords": [ "evolution", "gene networks", "fitness landscape learning" ], "content": "1. Introduction\n\nA central idea of modern biology is that evolution proceeds by mutation and selection. This process may be represented as a walk in a fitness landscape leading to fitness increase and slow adaptation 1 . According to classical ideas this walk can be considered a sequence of small random steps with small phenotypic effects. Nevertheless, there is a limited amount of experimental support for this idea 2 and some experimental evidence that evolution can involve genetic changes of a relatively large effect and that the total number of changes are surprisingly small 3 . Another intriguing fact is that organisms are capable of making adaptive predictions of environmental changes 4 .\n\nTo explain those facts new evolutionary concepts have been suggested (see the review by 5 and references therein). The main idea is that a population can “learn” (recognize) fitness landscapes 5– 7 . The approach developed in these works is a generalization of ideas from machine learning in which learning (regression to data) is viewed as selection and generalization (interpolation or extrapolation) is viewed as adaptation.\n\nA mathematical basis for investigation of evolution learning problems has been developed by 8. However, this work uses a simplified model, where organisms are represented as Boolean circuits seeking an “ideal answer” to environmental challenges. These circuits involve N g Boolean variables that can be interpreted as genes, and the ideal circuit answer maximizes the fitness. A similar model was studied numerically by 7 to confirm the theory of “facilitated variation” explaining the appearance of genetic variations which can lead to large phenotypic ones. In the work by 9 a theory of the evolution of these Boolean circuits was advanced. It was shown that, under some conditions—weak selection, see 10—a polynomially large population over polynomially many generations (polynomial in N g ) will end up almost surely consisting exclusively of assignments, which satisfy all constraints. This theorem can shed light on the problem of the evolution of complex adaptations since that satisfiability problem can be considered as a rough mathematical model of adaptation to many constraints.\n\nIn 6 it is shown that, in the regime of weak selection, population evolution can be described by the multiplicative weight update algorithm (MWUA), which is a powerful tool, well known in theoretical computer science and a generalization of such famous algorithms as Adaboost and others 11 . Note that in 6 infinitely large populations are investigated whereas the results of 9 hold only for finite populations and take into account genetic drift.\n\nEvolution Computation(EC) problems are considered recently by many papers 12– 16 mainly for artificial test fitness functions like OneMax or LeadingOnes (for an overwiew of EC problems, see 17).\n\nIn this paper, we investigate adaptation and the fitness landscape learning problem for more realistic fitness function This fitness can model adaptation for insects and connected with a fundamental hard combinatorial problem: K-SAT.\n\nThe main results can be outlined as follows. We show that, in a fixed environment, genes can serve as learners in the machine learning sense. Indeed, if an organism has survived for a long period, this fact alone constitutes important information, which can be used. The biological interpretation of this fact is simple: if a population is large enough and mutations are sufficiently rare, deleterious mutations are eliminated by purifying selection. Hence, those non-neutral mutant alleles which have become fixed in natural populations will, with probability close to 1, be adaptive and cause a positive increment of fitness (see Theorem 3.1 and Theorem 3.2 in Subsection 3.1 and Subsection 3.2). We obtain mathematical results, which allows us to estimate the reduction of mutation number due to that learning landscape procedure. Learning can sharply reduce the number of mutations needed to form a phenotypic trait useful for adaptation that is consistent with experimental data mentioned above (see 3.\n\nAnother important result is as follows. We estimate the accuracy of fundamental Nagylaki equations 6, 10 for a realistic population model, where the population size is bounded and a non-zero mutation rate is taken into account (in the case of asexual reproduction). Those accuracy estimates are fulfilled for all possible values of mutation rates and population sizes.\n\n\n2. Model\n\nIn this section, we describe our model and mathematical approach.\n\nWe assume that the genotype can be described by Boolean strings of length N g , where N g is the number of genes. Then\n\n\n\nwhere s i = 1 means that gene i is activated (switched on) and s i = 0 means that it is repressed (switched off). Correspondence between Boolean hypercube and genotypes is considered for example in 12.\n\nAlthough phenotype is controlled by genes, it is also influenced by environmental conditions and various epigenetic processes. In this paper, we suppose that phenotypic traits are controlled by genotype only. We consider levels f j of expressions of those traits as real variables in the interval (0, 1). Then the vector f = ( f 1, . . . , f N b ) can be considered to represent the organismal phenotype. We suppose that\n\n\n\nwhere f j ∈ (0, 1) is a real valued function of the Boolean string s, the genotype.\n\nOnly a part of s i is involved in f j . Namely, for each j we have a set of indices K j = { i 1, i 2, . . ., i n j } such that f j depends on s i with i ∈ K j , so that\n\n\n\nwhere i l ∈ K j and n j is the number of genes involved in the control of the trait expression. The representation of phenotype by the quantities f j is suggestive of quantitative traits because the f j are real valued. The limiting values of 0 or 1 suggest another interpretation, however, in terms of cell type. Multicellular organisms consist of cells of different types. One can suppose that the organismal phenotype is defined completely by the corresponding cell pattern. The cell type j is determined by morphogenes, which can be identified as gene products or signaling molecules that can change cell type (or genes that code for signaling molecules that can determine cell types or cell-cell interactions and then finally the cell pattern). The morphogene activity is defined by ( 2.2).\n\nWe further suppose:\n\nAssumption M. Assume activities f j have the following properties.\n\nThe sets K j are independent uniformly random subsets of S g = {1, . . . N g }\n\n\n\nWe denote the total number of genes involved in regulation of all f j by N r , where Nr=∑nj≤KNb.\n\nAssumption M implies that the genetic control of the phenotype is organized, in a sense, randomly, and that only a portion of the full set of genes controls phenotypic traits. That modularity of gene control is well known from experimental data (see 18, 19) and for evolution computation problems it was studied, for example, in 16.\n\nConsider an example, where the assumption M holds, where we have a saturated expression, inspired by earlier work 20, 21 . Let\n\n\n\nwhere j = 1, . . . , N b . Here σ( z) is a sigmoidal function of real z such that\n\n\n\nand w ij , h j are some coefficients (their meaning will be explained below). As an example, we can take σ( S) = (1 + exp( −bS)) −1, where b > 0 is a sharpness parameter. Note that for large b this sigmoidal function tends to the step function and for b = + ∞ our model becomes a Boolean one. The parameters h j defines thresholds for trait expression 20 . The relation ( 2.4) can be interpreted as a simple mathematical model for quantitative trait locus (QTL) action.\n\nTo understand the role of h j consider a trait f j and suppose that for a well adapted organism f j ≈ 1. Let, for simplicity, w ij take the values 1, 0, or −1. Then the parameter h j defines how many genes involved in the control of the f j expression should be activators and how many should be repressors. Let the numbers of activator and repressor genes be nj± respectively.\n\nThen f j ≈ 1 if nj+−nj−≫hj.\n\nOne can suppose that h j describes a direct influence of environment on phenotype, such as stress, that can exert epigenetic effects. In Section 2.5 using data from 3 we will show that the model defined by ( 2.4) are capable to describe main topological characteristics of really observed fitness functions in the case of mimicry, camouflage and thermoregulation for insects.\n\nLet us introduce the matrix W of size N b × N g with the entries w ij . The coefficients w ji determine the effects of terminal differentiation genes (see 23), and hence encodes the genotype-phenotype map. We assume that the coefficients w ji are random, with the probability that w ji > 0 or that w ji < 0 is β/2 N, where β > 0 is a parameter. This quantity β << N defines a genetic redundancy, i.e., averaged numbers of genes that control a trait. Note that then large β ≫ 1 one has n j < 4 β with the probability Pr β , which is exponentially close to 1: Pr β > 1 – exp(−0.1 β), thus, the number n j are bounded.\n\nWe know little about the details of how fitness relates to the phenotype of multicellular organisms, and for that reason classic neo-Darwinian theory takes fitness to be a function of genotype. Some models which take account of epistasis have been proposed 24 . The random field models assign fitness values to genotypes independently from a fixed probability distribution. They are close to mutation selection models introduced by 25, and can be named House of Cards (HoC) model. The best known model of this kind is the NK model introduced by Kauffman and Weinberger 26 , where each locus interacts with K other loci. Rough Mount Fuji (RMF) models are obtained by combining a random HoC landscape with an additive landscape models 27 . In evolution computations (EC) some artificial fitness models were used, for example OneMax and Leading Ones to test evolution algorithms, see for example 15 .\n\nIn this work, we use the classical approach of R. Fisher by introducing an explicit representation of phenotype, f, and allow it to determine fitness through interaction with an environment b. That is, we assume that the phenotype is completely determined by the phenotype trait expression, and thus the fitness depends on the genotype s via f j .\n\nWe express the relative fitness F and its dependence on environment b via an auxiliary function W via the relation\n\n\n\nwhere K F is a positive constant and b = ( b 1,..., b N b ) is a vector consisting of coefficients b j , respectively. Below we refer to W as a fitness potential, and we assume that\n\n\n\nSometimes, if the parameter b is fixed, we shall omit the corresponding argument in notation for W and F.\n\nWe consider fitness as a numerical measure of interactions between the phenotype and an environment. For a fixed environment, this idea gives us the fitness of classical population genetics. A part of the fitness, however, depends on the organism developing properly and for now we represent it as independent of the environment, although we are aware that this is not always the case. Note that some coefficients b j may be negative and others may be positive, and that the model ( 2.7) can describe gene epistatic effects via dependence of f j on s if f j are nonlinear in s.\n\nThe expression ( 2.7) can serve as a rough approximation of the fitness function in the case of insects such as grasshoppers or fruit flies. In fact, important factors, which determine insect survival, are thermoregulation, mimicry and camouflage levels 18, 22, 28 . All those factors depend on colour pigmentation pattern. Blackwhite pigmentation patterns can be roughly described by vectors f = ( f 1, f 2,..., f N b ), where f j ≈ 1 and f j ≈ 0 mean that the cell j is black, or white, respectively, Then thermoregulation depends on ∑jfj. The mimicry level can be approximately defined by expression ∑j\|fj−fj\| , where f is a target pattern corresponding to an insect to mimic. Colour patterns can be also described by classical RGB formalism.\n\nThe representation of the fitness as a sum of terms ( 2.7) is of course a rough approximation; however if assumption M holds that representation is consistent with important observed facts. First, mutations have been identified that alter one part of the pigment pattern without affecting any other. This independence of different pattern parts can be explained by the modular organization of the genetic regulation that controls pigmentation. In the course of evolution, different aspects of the pigment pattern have clearly evolved independently of each other 18 . Second, the topology of the fitness landscapes was studied in 22 by field experiments in the case of insect mimicry. Main conclusions are as follows. A number of studies of fitness landscapes in natural populations have demonstrated low fitness of intermediate phenotypes, i.e., existence of valley in the fitness landscape. It is found 22 that natural selection promotes genetic architecture preventing the expression of intermediate phenotypes. Close fitness peaks are separated by ridges, favouring colour pattern switches and allowing drift from local peaks.\n\nIn Section 2.5 we will show that the fitness model defined by ( 2.4) and ( 2.7) have those topological properties.\n\nFor simplicity, we consider populations with asexual reproduction. (Although a part of the results remain valid for sexual reproduction, as we discuss at the end of this subsection). We choose initial genotypes randomly from a gene pool and assign them to organisms. This choice is invariant with respect to the population member, i.e,. the probability to assign a given genotype s to a member of the population does not depend on that member.\n\nIn each generation, there are N pop( t) individuals, the genome of each of which is denoted by s( t), where t = 0, 1, 2,... stands for the evolution step number). Following the classical Wright-Fisher ideas, we suppose that generations do not overlap. In each generation (i.e., for each t), the following three steps are performed:\n\n1. Each individual s at each evolution step can mutate with probability p mut per gene;\n\n2. At evolution step t each individual with a genotype s produces k progeny, where k is a random non-negative integer, distributed according to the Poisson law\n\n\n\nwhere q = F( s) is the fitness of that individual;\n\n3. To take into account ecological restrictions on the population size, we introduce the maximal population size N popmax . If N′ ( t) > N popmax, where N′ ( t) is the number of progeny produced by the population at step t, we kill randomly selected individuals in a population-dependent manner. The probability of the death of an individual is given by p kill( N′ ) = 1 − ( N popmax/ N′ ( t)). If N′ ( t) ≤ N popmax, we do nothing. We refer to this as the “massacre procedure.”\n\nConditions 1 and 2 imply that mutations in the genotypes create a new genetic pool and then a new round of selection starts. Condition 3 expresses the fundamental ecological limitation that all environments can only support populations of a limited size. If N popmax ≫ 1 then by ( 2.8) and the Central Limit Theorem one can show, under some additional conditions, (see Section 4) that fluctuations of the population size are small, and thus the population is ecologically stable and N pop( t) ≈ N popmax.\n\nIn the limit case of infinitely large populations we will write the discrete dynamical equation for the time evolution of the frequency X( s, t) of the genotype s in the population as\n\n\n\nwhere F¯ ( t) is the average fitness of the population at the moment t defined by\n\n\n\nwhere S( t) is the set of genotypes existing in the population at time t (the genetic pool) and X( s, t) = N( s, t) /N pop( t) is the frequency of the genotype s. Here N( s, t) denotes the number of the population members with the genotype s at the step t.\n\nThe equations ( 2.9) do not take mutations into account. They describe only changes in the genotype frequencies because of selection at the t-th time step. The same equations govern evolution in the case of sexual reproduction in the limit of weak selection 6, 10 . Note that for an evolution defined by ( 2.9), the average fitness F¯ ( t) defined by ( 2.10) satisfies Fisher’s theorem, so that this function increases at each time step t: F¯ ( t + 1) ≥ F¯ ( t).\n\nAdaptation (i.e., maximization of fitness in a changing environment) is a very hard problem since over evolutionary history we observe the coevolution of many traits accompanied by changes in many genes. In its general context, this is a problem in the theory of macroevolution, which in general requires the integration of population genetics and developmental biology for its full understanding. There are two key components of this problem. First, development is itself a dynamical process operating over time. Second, there is a combinatorial component of development wherein different combinations of gene must be expressed in different cell types. This combinatorial aspect of the problem means that straightforward theoretical methods of considering the relationship between gene expression and a changing environment that have been very successful in single celled organisms 29 cannot be applied to metazoa. In this work, for the sake of tractability, we focus on the combinatorial aspect of the problem and neglect developmental dynamics. Even at the highly simplified level of our model, adaptation is a hard computational problem, as we now demonstrate.\n\nConsider the case, where f j are defined by relations ( 2.4) and assume that\n\ni) σ is the step function;\n\nii) b j > 0.\n\nAs a consequence of the second assumption, F attains its maximum for f 1 = 1, f 2 = 1,..., f N b = 1. Let us show that, even in this particular case, the problem of the fitness maximization with respect to s is very complex. In fact, for a choice of h j it reduces to the famous NP-complete problem, so-called K-SAT, which has received a great deal of attention in the last few decades (see 30– 35). The K-SAT can be formulated as follows.\n\nK -SAT problem. Let us consider the set V n = { s 1,..., s n } of Boolean variables s i ∈ {0, 1} and a set m of m clauses. The clauses C j are disjunctions (logical ORs) involving K literals z i 1 , z i 2 ,..., z i k , where each z i is either s i or the negation s¯ i of s i. The problem is to test whether one can satisfy all of the clauses by an assignment of Boolean variables.\n\nCook and Levin 30, 31 have shown that the K-SAT problem is NP-complete and therefore in general it is not feasible in a reasonable running time. In subsequent studies—for instance, by 32 —it was shown that K-SAT of a random structure is feasible under the condition that N b < α c ( K) N g , where α c ( K) ≈ 2 K log 2 for large K.\n\nThe set K of solutions of random K-SAT has a nontrivial structure depending on parameter α = N b/N g 33, 35 . For sufficiently small α < α g ( K), where α g ( K) ≈ 2 K log( K) /K is some critical parameter, the set K forms a giant cluster, where nearest solutions are connected by a single flip and one can go from a solution to another by a sequence of single flips (pointed mutations) 35 . For α ∈ ( α g , α d ), where α d ( k) < α c is another critical value, solutions form a set of disconnected clusters. The local search algorithms do not work in the domain α > α g .\n\nProbably, for evolution context K-SAT was applied first in 36, where it was used for an investigation of speciation problem.\n\nTo see the connection of our model with K-SAT, consider equation ( 2.4) supposing that w ij ∈ {1, 0, −1} and h j = −C j + 0.5, where C j is the number of negative w ji in the sum Sj=∑i=1Ngwjisi. We set m = N b and n = N g . Under this choice of h j , the terms σ( S j ) can be represented as disjunctions of literals z j . Each literal z j equals either s j or s¯ j , where s¯ j denotes negation of s j . To maximize the fitness, we must assign s j such that all disjunctions will be satisfied. If we fix the number n j of the literals participating in each disjunction (clause) and set n j = K, this assignment problem is precisely the K-SAT problem formulated above.\n\nReduction to the K-SAT problem is a transparent way of representing the idea that multiple constraints need to be satisfied. The number K defines the gene redundancy and the probability of gene pleiotropy. Remind that pleiotropy occurs when one gene influences two or more seemingly unrelated phenotypic traits. The threshold h j and K define the number of genes which need be flipped in order to attain a high expression of the trait f j . Note that gene pleiotropy is a fundamental characteristics 37 , which is studied for real organisms only recently (see 19). We can compare experimental observations and consequence of model ( 2.4), which is a generalisation of K-SAT (compare plot Figure 1 and plots on Figure 1 in 19). So, we can fit our model parameters using real data. Moreover, we can check validity of our model by the following arguments.\n\nWe note that, in the case of giant cluster formation, the topological properties of the solution set K , mentioned above, outline the properties of really observed fitness landscapes 22 : existence of many peaks, valleys and ridges connecting peaks. Namely, existence of many solutions of K-SAT, when a giant cluster exists, means that the landscape has a number of peaks separated by valleys. On the other hand, connectance of solutions within the giant cluster can be interpreted that there exist ridges that connect peaks.\n\nNote that there are important differences between K-SAT in Theoretical Computer Science and fitness maximization problems. First, the signs of b j are unknown for real biological situations since the fitness landscape is unknown. Second, our adaptation problem involves the threshold parameters h j (see ( 2.4)). In contrast to K-SAT, in our case the Boolean circuit is plastic, because the h j are not fixed.\n\nIf the b j are unknown, the adaptation (fitness maximization) problem becomes even harder because we do not know the function to optimize. Therefore, many algorithms for K-SAT are useless for biological adaptation problems. Below we will nonetheless obtain some analytical results based on the assumption that b j are random.\n\n\n3. Main theorems\n\nThe subsequent material is organized as follows. First we formulate a result on regulation mechanism power. Furthermore, we prove two fitness landscape learning theorems.\n\nFor simplicity, we consider asexual reproduction. To obtain similar results for sexual reproduction, one can consider a weak selection regime and use the results of 10, where eq. ( 2.9) are derived.\n\nLet us introduce two sets of indices I + and I − , such that I + ∪ I − = {1,..., N b }. We refer to these sets in the sequel as positive and negative sets, respectively. We have\n\n\n\n\n\nThe biological interpretation of that definition is transparent: the expression of the traits f j with j ∊ I + increases the fitness and for j ∊ I – expression of the trait decreases the fitness.\n\nLet s and s¯ be two genotypes. Then we denote by Diff( s, s¯ ) the set of positions i such that s i ≠ s¯ i :\n\n\n\nThe set Diff( s, s¯ ) indicates which genes in s should be flipped in order to obtain s¯ .\n\nWe formulate two theorems on fitness landscape learning. First we consider the case of infinitely large populations.\n\nTheorem 3.1. Suppose that the evolution of the genotype frequencies X( s, t) is determined by equations ( 2.9) and ( 2.10). Moreover, assume that\n\nI for all t ∊ [ T 1, T 1 + T c ] , where T 1, T c > 0 are integers, the population contains two genotypes s and s¯ such that the frequencies X( s, t) and X( s¯ , t) satisfy\n\n\n\nII we have\n\n\n\nfor some j. In other words, the genes s i such that s i ≠ s¯ i are involved in a single regulation set K j ; and finally,\n\nIII Let\n\n\n\nand\n\n\n\nThen, if\n\n\n\nwe have j ∊ I +. If f j ( s) > f j ( s¯ ), then j ∊ I – .\n\nBefore proving this, let us make some comments. The biological meaning of the theorem is simple: for simple fitness models, where unknown parameters b j are involved in a linear way, in the limit of infinitely large populations fitness landscape learning is possible.\n\nMoreover, note that we do not make any specific assumptions about the nature of mutation, but only that all genetic variation between s and s¯ are contained in a single regulatory set K j .\n\nThe assertion of Theorem 3.1 is not valid if the set Diff( s, s¯ ) belongs to a union of different regulation sets K j , j = j 1, . . . , j p with p > 1. This effect of belonging to different sets K j is pleiotropy in gene regulation. Note that if N b ≪ N g then the pleiotropy probability is small for large genome lengths N g . On the contrary, if N b ≫ N g then assumption II is invalid.\n\nAssumption II looks natural if when we deal with point mutations. In fact, if s¯ is obtained from s by a single point mutation then condition ( 3.4) always holds for some j. For small mutation rates the probability of two point mutations is essentially below than the probability of a single mutation.\n\nTo conclude let us note that Theorem gives a rough estimate for the learning time T c :\n\n\n\nProof. The main idea is simple. Negative mutations lead to elimination of mutant genotypes from the population, and the corresponding frequencies become, for large times, exponentially small.\n\nAssume that ( 3.7) holds. Let j ∊ I − , and thus b j < 0. Consider the quantity\n\n\n\nAccording to assumption II\n\n\n\nAssumption III entails that\n\n\n\nRelations ( 2.6) and ( 3.10) imply\n\n\n\nBy ( 2.9) and the last inequality we find that for T > T 1\n\n\n\nConsider inequality ( 3.11) for T = T 1 + T c . Let us note that in the relation Q( T 1) = X( s, T 1) /X( s¯ , T 1) the numerator is p 0 whereas the denominator ≤ 1. Thus, Q( T 1) ≥ p 0. The same arguments show that Q( T 1 + T c ) ≤ 1 /p 1. Therefore, by ( 3.11) one obtains that\n\n\n\nThis inequality leads to a contradiction for T c satisfying ( 3.6), thus completing the proof.\n\nTheorem 3.1 can be extended to the case of finite populations and non-zero mutation rates. is small. To formulate this generalization, we need an additional assumption about the fitness function. Suppose that\n\n\n\n\n\nwhere c F , C F > 0 are constants independent of t. For example, if\n\n\n\nthen c F = K F exp(– γ) and C F = K F exp( γ) and ( 3.13) holds if K F > exp( γ).\n\nCondition ( 3.13) means that each individual gives birth to at least c F and at most C F descendants, where those bounds do not depend on the population size and evolution step.\n\nLet\n\n\n\nNote that for simplicity in the next Theorem 3.2 we consider point mutations (bit flipping) only. The model used here cannot represent mutations of arbitrarily small effect, but it can include insertions or deletions. In contrast to Theorem 3.2, Theorem 3.1 is valid for all kinds of mutations.\n\nThen we have\n\nTheorem 3.2. Consider the population dynamics defined by model 1 - 3 in Subsection 2.4. Assume conditions ( 3.14) and M hold, and assumptions ( 3.3), ( 3.4), ( 3.5), ( 3.7) of Theorem 3.1 are satisfied. Suppose\n\n\n\n\n\nThen if j ∊ I – the inequality\n\n\n\nis fulfilled with the probability Pr v such that\n\n\n\nwhere for large N popmax and p mut → 0\n\n\n\nInterpretation of Theorem 3.2\n\nIt is interesting to compare Theorem3.1 and Theorem3.2. The previous one asserts that for infinite populations the probability of the event j ∊ I − is zero whereas the second one claims that this probability becomes exponentially small as the population size increases.\n\nThis theorem also shows that evolution can make a statistical test checking the hypothesis H – that j ∊ I – against the hypothesis H + that j ∊ I +. Suppose that H – is true. Let V be the event that the frequency X( s¯ , t) of the genotype s¯ in the population is larger than p 1 within a sufficiently large time T c . According to estimate ( 3.18), the probability of the event V is so small that it is almost unbelievable. Therefore, the hypothesis H – should be rejected. We will refer T c as the checking time.\n\nIdeas for the proof. The main idea is the same as that for the previous theorem: we compare the frequencies of the organisms with the genotype s¯ and the organisms with the genotype s. However, the proof includes a number of technical details connected with estimates of mutation effects and fluctuations. The formal proof can be found in Section 4. It is based on estimates of the accuracy of the Nagylaki equations (in the case of asexual reproduction).\n\n\n4 Proof of theorems\n\nLet us prove Theorem 3.2.\n\nLet us introduce notation and make some preliminary remarks. Remind that we denote by N ( s, t) the number of the population members with the genotype s at the moment t. Let X( t) be the set of all population members at the moment t. For each x ∊ X( t) let us denote by N′ ( x, t) the number of progeny born by the individual x at the moment t before the massacre (see point 3 of model from Subsection 2.4). Let s g ( x) be the genotype of x. Then, according to ( 2.8), the mean of N′ ( x, t) is\n\n\n\nwhere EX denotes the expected value of X. By N− ( s, t) we denote the number of all progeny born by individuals with the genotype s at the moment t before the massacre. Since all progeny are produced independently and randomly, the previous relation gives\n\n\n\nOur main analytical tools are the Chernoff bounds and the Hoeffding inequalities. We also use the Markov inequality: for a positive random quantity X and a > 0 one has\n\n\n\nMoreover, we use two elementary estimates. Let be an event in stochastic population dynamics. We denote by Not the negation (complement) of and by Pr( \|ℬ) the conditional probability of under the condition ℬ. For events , ℬ 1, . . . , ℬ n we have\n\n\n\nFor two events , ℬ one has\n\n\n\nLemma 4.1. Let X i be independent random quantities, where i = 1, . . . , n. Let each X i be distributed according to the Poisson law with the average EX i = μ i. Let us denote\n\n\n\nThen for all δ > 0\n\n\n\nwhere\n\n\n\nSimilarly,\n\n\n\nProof. Note that for any λ > 0\n\n\n\nSince X j are independent quantities, we have\n\n\n\nThe straight forward computation shows that\n\n\n\nTherefore, due to the Markov inequality ( 4.3) and estimate ( 4.8) one has\n\n\n\nwhere\n\n\n\nWe minimize f with respect to λ and obtain ( 4.6). To derive ( 4.7), we use\n\n\n\nand repeat the same arguments. The Lemma is proved.\n\nLemma 4.2. Let X i be independent random quantities, where i = 1, . . . , n such that X i ∊ {0, 1} and EX i = p. Then\n\n\n\nwhere\n\n\n\nProof. Note that for any λ > 0\n\n\n\nSince X j are independent quantities, we have\n\n\n\nNote that E exp( −λX j ) = p exp( −λ) + 1 − p. Let\n\n\n\nWe take λ = ln 2 and find that G(ln 2, p) = −g( p). Now by using the Markov inequality ( 4.3) and estimate ( 4.12) one obtains ( 4.10). The Lemma is proved.\n\nWe also use the following Chernoff-Hoeffding theorem. Let X i be i.i.d. quantities such that X i ∊ {0, 1} and EX i = p, where i = 1, . . . , n. Then for X=∑j=1nXj one has\n\n\n\nwhere D( x\|\|y) is the Kullback-Leibler divergence\n\n\n\nMoreover, we will use the Hoeffding Theorem: if i.i.d. quantities X i ∊ [0, 1] with the probability 1 then\n\n\n\nFirst we estimate the population size fluctuations.\n\nLemma 4.3. Let N¯ ( t) be the number of all progeny, born in the population at the moment t before the massacre, and ε 1 > 0 be a small number. Then\n\n\n\nwith probability\n\n\n\nwhere\n\n\n\nProof. Let n′ ( x, t) denote the number of progeny produced by the individual x before the massacre at the t-th evolution step. The number N¯ ( t) is the sum\n\n\n\nof the mutually independent random quantities. According to ( 4.2), the average E N′ ( x, t) is F( s g ( x)). Therefore,\n\n\n\n\n\n\n\nWe set\n\n\n\nand use the Lemma 4.1 that gives us ( 4.17).\n\nLemma 4.4. Let ε 2 ∊ (0, 1) be fixed and condition ( 3.13) be fulfilled. Assume, moreover, that\n\n\n\nwhere\n\n\n\nand c F > 1 is defined by ( 3.13). Let us define the event ε 2 ( t) by\n\n\n\nThen one has\n\n\n\nwhere\n\n\n\nand\n\n\n\nProof. Let ξ( x) be random quantities defined as follows: ξ( x) = 1 if the individual x is survived as a result of massacre (see point 3 of our model from Subsection 2.4), and ξ( x) = 0 otherwise. Let X′ ( t) be the set of progeny produced by all individuals from the population. Then the number N sur( t) = N pop( t + 1) of finally survived progeny can be computed as follows:\n\n\n\nNote that \| X′ ( t) \| = N¯ ( t). Moreover, Eξ( x ) = N popmax/ N¯ ( t) for N¯ ( t) ≥ N popmax. Therefore, if N¯ ( t) ≥ N popmax then\n\n\n\nLet us define the event\n\n\n\nwhere the interval J ε ( t) is defined by ( 4.16) and ε˜ is defined by ( 4.27). By ( 4.4) we have\n\n\n\nNow we apply the Hoeffding inequality ( 4.15). For each ε 2 > 0 we obtain\n\n\n\nIf ℬ ( t) takes place, then N¯ ( t) ≥ N popmax and consequently\n\n\n\nTherefore,\n\n\n\nMoreover, by Lemma 4.3\n\n\n\nInequalities ( 4.30), ( 4.32) and ( 4.33) prove ( 4.25).\n\nThe following lemma, in particular, allows us to obtain equations ( 2.9) and ( 2.10) in the limit of infinite populations and for small mutation probabilities.\n\nRecall that N¯(s,t) denotes the number of non-mutated progeny generated by the individuals with the genotype s before the massacre. Let N sur( s, t) be the number of those progeny that survived after that massacre.\n\nLemma 4.5. Let ε 0 be a positive number satisfying ( 4.75) and\n\n\n\nThen one has\n\n\n\nwith the probability Pr s, t,+ such that\n\n\n\nwhere\n\n\n\n\n\n\n\n\n\n\n\nSimilarly,\n\n\n\nwith the probability Pr s, t,− such that\n\n\n\nProof. Step 1, estimates of fluctuations. First let us estimate the fluctuations of the number N¯(s,t) . For each ε 2 > 0 let us define the event\n\n\n\nBy Lemma 4.1 one has\n\n\n\nNote that\n\n\n\nAs a result, by ( 4.46) we obtain\n\n\n\nStep 2. Here we estimate the number of progeny that survived as a result of the massacre procedure (point 3 of the population dynamics model, see subsection 2.9). Let X' ( s, t) be the set of progeny produced by individuals with the genotype s. Then the number N sur( s, t) of survived progeny x for individuals x belonging to the set Z' ( s, t) is\n\n\n\nwhere ξ( x) are defined in the proof of the previous Lemma. For ε 3 > 0 we consider the event\n\n\n\nLet us estimate the probability Pr( sur, s ( t)). According to the Hoeffding Theorem ( 4.15)\n\n\n\nNote that ξ( x) and ξ( y) are independent quantities for different x and y, thus under the condition N¯(t) > N popmax\n\n\n\ntherefore,\n\n\n\nLet us define the events ℬ s ( t) and ℬ( t) by\n\n\n\n\n\nThen using ( 4.4) one has\n\n\n\nWe observe that under conditions ℬ s ( t) and ℬ( t)\n\n\n\nIn that estimate let us set ε 2 = 0.5 and ε 1 = 1. Taking into account that E N¯(t) = F¯(t) N pop( t) < 2C FN popmax, we have that\n\n\n\nwhere\n\n\n\nMoreover, according to ( 4.47)\n\n\n\nand due to ( 4.17)\n\n\n\nwhere R 1, R 2 are defined by ( 4.37) and ( 4.38). Finally,\n\n\n\nStep 3, estimate of the number of mutants.\n\nLet us estimate how many individuals with genotypes s can mutate. The probability of mutation is p mut. Let N mut( s, t) be the number of such mutants. Let us define the event mut, s ( t) by\n\n\n\nSince the random quantity N mut( s, t) is subject to the Bernoulli law, we can apply the Chernoff-Hoeffding inequality ( 4.13). Then we obtain that\n\n\n\nwhere, according to definition ( 4.14) of D( x\|\| y), one has\n\n\n\nand g is defined by ( 4.11).\n\nUsing ( 4.4) one has\n\n\n\nAs a result, by Lemma 4.3 one finds\n\n\n\nwhere R 4, R 5 are defined by ( 4.40) and ( 4.41).\n\nTo prove ( 4.35), we set ε 3 = ε 0/2. Taking into account condition ( 4.75) for ε 0 we see that if the both events Not mut, s ( t) and Not sur, s, ε 0/2( t) take place, then inequality ( 4.35) is fulfilled. Thus\n\n\n\nwhere R i are defined by ( 4.37)–( 4.41).\n\nFinally, taking into account the results of steps 1, 2 and 3 we see that estimate ( 4.35) holds with the probability Pr t,+. It completes the proof of ( 4.35). The second inequality ( 4.42) can be obtained in the same way.\n\nWe use the same idea that in the proof of Theorem 3.1 but first we establish uniform bounds for the population size and other quantities involved in the proof.\n\nStep 1 Here we estimate the population size. Let us set\n\n\n\nin Lemma 4.4. Let us consider the events ε 2 ( t) defined by ( 4.24) in Lemma 4.4. If the events ε 2 ( t) take place for all t ∈ [ T 1, T 1 + T c ] and N pop(0) = N popmax, we have that\n\n\n\n\n\nThen conditions ( 3.15), ( 3.16) of Theorem 3.2 imply\n\n\n\nThose inequalities imply the following estimates for the quantities R i defined by ( 4.37) –( 4.41):\n\n\n\nwhere q i are defined by\n\n\n\n\n\n\n\n\n\n\n\nwhere\n\n\n\nand\n\n\n\nwhere ε˜ = 1 – ( κc F ) –1, and\n\n\n\nFor each p ∈ (0, 1) let us define an auxiliary function\n\n\n\nwhere q i , q˜ are defined by relations ( 4.68)–( 4.72). We can find asymptotics of ρ under natural assumptions that p mut → 0 and N popmax → ∞ while all the rest parameters are fixed. Then the leading term in the right hand side of ( 4.36) is q 4 and U(pmut)=(2ln⁡2−1)pmut+O(pmut2). As a result, we have\n\n\n\nStep 2. Let Q( t) is defined by ( 3.8) and, moreover, let j ∈ I −. We use Lemma 4.5 inductively for genotypes s and s˜ . Let us set\n\n\n\nwhere ε is defined by ( 4.75). We remark that the inequality\n\n\n\nholds with a probability Pr Q, t > 0. Let us obtain a uniform estimate of that probability. Let ℰ( t) be the event that ( 4.78) holds at the step t. Using ( 4.4) we have\n\n\n\nwhere, according to Lemma 4.4, the probability of the event Not ε 2 ( t) is less than η, where η is defined by ( 4.25), and\n\n\n\nWe conclude by ( 4.5) that\n\n\n\nwhere q˜ is defined by ( 4.74). This estimate is uniform in t ∈ [1, . . . , T c ]. By ( 4.80) we obtain then that the inequality\n\n\n\nis satisfied with the probability Pr v such that\n\n\n\nFor ε 0 defined by ( 4.75). one has\n\n\n\nNow repeating the same arguments that in the end of the proof of Theorem 3.1, and taking into account asymptotics ( 4.77), we obtain the conclusion of Theorem 3.2.\n\n\n5 Discussion\n\nIn this paper, we proposed a model for fitness landscape learning, which extends earlier work by 7– 9 in two ways. First, we use hybrid circuits involving two kinds of variables. The first class of variables are real valued in the interval (0, 1) and can be interpreted as relative levels of phenotypic traits, other variables are Boolean and can be interpreted as genes. Second, we use a threshold scheme of regulation, which is inspired by ideas of the paper by 20. All variables are involved in gene regulation via thresholds.\n\nThe work presented here is a major extension of a long term effort to explicitly model the effects of phenotypic buffering in evolution by considering a class of Boolean and mixed Boolean-continuous models in which the phenotype is represented explicitly and the degree of phenotypic buffering can be controlled in various ways. For example, we have demonstrated that the idea of an “evolutionary capacitor” 38, 39 can be implemented by explicit control of phenotypic buffering in a hub-and-spokes architecture 23 and that in a more general class of genetic architecture numerical simulations show that an intermediate level of buffering is optimal for evolution in a changing environment 40 .\n\nThe results reported here are very promising, since they are consistent with the results of recent experiments by 41 and 42 on heat shock stress. The essential mechanism is that the exploration of the fitness landscape by the genetic network in such a way that future mutations are more likely to be adaptive. We have shown that, at least for some fitness landscapes, rapid evolutionary changes—perhaps instances of the “hopeful monsters” of Goldschmidt 43 —can be created by a combination of random small mutations and epigenetic effects. The main idea is that small mutation pave the way for large epigenetic or genetic changes. The hypothetical mechanism, which we propose, can be outlined as follows (see Figure 2, Figure 3).\n\nAt initial moment the trait expressions take the values x = 0.5, y = 0.5. According to Fisher’s ideas, random large mutations decrease the fitness F = K F exp( W). (Changes of x = F 1, y = f 2, which are induced by mutations, are shown by red vectors.) Thus such mutations produce non-viable organisms.\n\nAt the initial time the trait expressions take the values x = 0.5, y = 0.5. Evolutionary changes go in two stages. First we make small random mutations (shown by red vectors), which explore the fitness landscape. If such a mutation is not eliminated from the population, this means that a correct evolution direction is found, and gene regulation system makes a big leap (shown by the green vector) in the direction of that small mutation. Such a two step model can be called clever Goldshmidt leaps. Note that evolution is gradual, and the existence of clusters of almost identical genes involved in the same QTL increases the chances to create a clever Goldschmidt hopeful monster.\n\nThe expression of genes involved in the expression of phenotypic traits depends on threshold parameters h j , which take three values: a large negative one, a neutral value close to zero and a large positive one. First the threshold parameter h j is small and thus the phenotypic trait is sensitive with respect to even small mutations. Those mutations play a fundamental role working as scouts exploring environments (see Figure 3). If a mutation occurred and the corresponding mutant has survived within T c ≫ 1 generations then according to Theorem 3.1 and Theorem 3.2 these events mean that that mutation increases the fitness that allows the network to estimate the correct direction of evolution. Then gene regulation detects that increase to change the threshold according to simple rules. Namely, if the trait is less expressed in that mutant with respect to wild type parent, the gene regulation system decreases the threshold up to the large negative value. On the contrary, if the trait is strongly expressed in the mutant, the gene regulation system increases the threshold up to the large positive value. This simple regulation control not only sharply reduces the number of mutations needed for adaptation, but also canalizes the phenotype since for large thresholds the trait expression level becomes insensitive with respect to mutations. We suppose that these threshold modifications can be inherited.\n\nSo, we propose the mechanism: small mutations serve as scouts finding the way for large epigenetic or genetic changes, which can be performed by gene regulatory system.\n\nThe mechanism may also explain the results of 4 on prediction of environmental changes. In fact, let us suppose that environment varies in time. The first, perhaps relatively small, variations can trigger the threshold mechanism described above. As a result, the population will be adapted to the subsequent changes in advance.\n\nOur results show that evolution can proceed rapidly because it reduces the number of mutations required for adaptive change.\n\nThe primary limitation of our results is that the representation of the evolving genetic network is limited to the network of gene controlling phenotype, represented here by the Boolean strings s. Other model variables represent the coarse-grained activities of genes. One class is the terminal differentiation genes represented by w ij , and another are the genes or epigenetic factors controlling the thresholds h and their associated learning rules. A more careful consideration of the relationship of these moieties to observable molecular entities is an important objective of future work. At the mathematical level, the key analytical results were obtained in a simplified context that falls short of a realistic level of pleiotropy and thus of the level of NP-hard complexity exhibited by fully pleotropic forms of our model. We believe that our analytical results can be generalized, which we plan to address in future work.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe second author was supported by the grant of Russian Ministry of Education, 2012-1.2.1-12-000-1013-016. Additionally, the second author was financially supported by Government of Russian Federation, Grant Grant 08-08. D. Grigoriev is grateful to the grant RSF 16-11-10075 and to both MCCME and MPI f\\\"ur Mathematik for wonderful working conditions and inspiring atmosphere. J. Reinitz and S. Vakulenko were supported by US NIH grant RO1 OD010936 (formerly RO1 RR07801).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nOrr HA: The genetic theory of adaptation: a brief history. Nat Rev Genet. 2005;6(2):119–127. PubMed Abstract \| Publisher Full Text\n\nOrr HA, Coyne JA: The genetics of adaptation: a reassessment. Am Nat. 1992;140(5):725–742. PubMed Abstract \| Publisher Full Text\n\nZeyl C: The number of mutations selected during adaptation in a laboratory population of Saccharomyces cerevisiae. Genetics. 2005;169(4):1825–1831. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMitchell A, Romano GH, Groisman B, et al.: Adaptive prediction of environmental changes by microorganisms. Nature. 2009;460(7252):220–4. PubMed Abstract \| Publisher Full Text\n\nWatson RA, Szathmáry E: How Can Evolution Learn? Trends Ecol Evol. 2016;31(2):147–157. PubMed Abstract \| Publisher Full Text\n\nChastain E, Livnat A, Papadimitriou C, et al.: Algorithms, games, and evolution. Proc Natl Acad Sci U S A. 2014;111(29):10620–10623. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nParter M, Kashtan N, Alon U: Facilitated variation: how evolution learns from past environments to generalize to new environments. PLoS Comput Biol. 2008;4(11):e1000206. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nValiant LG: Evolvability. Journal of the ACM. 2009;56(1):1–21. Publisher Full Text\n\nLivnat A, Papadimitriou C, Rubinstein A, et al.: Satisfiability and evolution. In Annu IEEE Symp Found Comput Sci FOCS.2014;524–530. Publisher Full Text\n\nNagylaki T: The evolution of multilocus systems under weak selection. Genetics. 1993;134(2):627–647. PubMed Abstract \| Free Full Text\n\nArora S, Hazan E, Kale S: The Multiplicative Weights Update Method: A Meta-Algorithm and Applications. Theory Comput. 2012;8:121–164. Publisher Full Text\n\nPaixão T, Badkobeh G, Barton N, et al.: Toward a unifying framework for evolutionary processes. J Theor Biol. 2015;383:28–43. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMühlenbein H, Schlierkamp-Voosen D: Predictive Models for the Breeder Genetic Algorithm I. Continuous Parameter Optimization. Evol Comput. 1993;1(1):25–49. Publisher Full Text\n\nChatterjee K, Pavlogiannis A, Adlam B, et al.: The time scale of evolutionary innovation. PLoS Comput Biol. 2014;10(9):e1003818. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHeredia JP, Trubenová B, Sudholt D, et al.: Selection Limits to Adaptive Walks on Correlated Landscapes. Genetics. 2017;205(2):803–825. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDoerr B, Doerr C, Kötzing T: Solving Problems with Unknown Solution Length at (Almost) No Extra Cost. In Proceedings of the 2015 on Genetic and Evolutionary Computation Conference GECCO - ’15. Madrid, Spain, ACM Press,2015;831–838. Publisher Full Text\n\nRudolph G: Finite Markov Chain Results in Evolutionary Computation: A Tour d’Horizon. Fundamenta Informaticae. 1998;35(1–4):67–89. Publisher Full Text\n\nWitcopp PJ, Carroll SB, Kopp A: Evolution in black and white: genetic control of pigment patterns in Drosophila. Trends Genet. 2003;19(9):495–504. PubMed Abstract \| Publisher Full Text\n\nWang Z, Liao BY, Zhang J: Genomic patterns of pleiotropy and the evolution of complexity. Proc Natl Acad Sci U S A. 2010;107(42):18034–18039. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStern C: Selection for subthreshold differences and the origin of pseudoexogenous adaptations. Am Nat. 1958;92(866):313–316. Publisher Full Text\n\nMjolsness E, Sharp DH, Reinitz J: A connectionist model of development. J Theor Biol. 1991;152(4):429–453. PubMed Abstract \| Publisher Full Text\n\nArias M, le Poul Y, Chouteau M, et al.: Crossing fitness valleys: empirical estimation of a fitness landscape associated with polymorphic mimicry. Proc Biol Sci. 2016;283(1829): pii: 20160391. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGrigoriev D, Reinitz J, Vakulenko S, et al.: Punctuated evolution and robustness in morphogenesis. Biosystems. 2014;123:106–113. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFranke J, Klözer A, de Visser JA, et al.: Evolutionary accessibility of mutational pathways. PLoS Comput. Biol. 2011;7(8): e1002134. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKingman JFC: A simple model for the balance between selection and mutation. J Appl Probab. 1978;15(1):1–12. Publisher Full Text\n\nKauffman SA, Weinberger ED: The NK model of rugged fitness landscapes and its application to maturation of the immune response. J Theor Biol. 1989;141(2):211–245. PubMed Abstract \| Publisher Full Text\n\nAita T, Iwakura M, Husimi Y: A cross-section of the fitness landscape of dihydrofolate reductase. Protein Eng. 2001;14(9):633–638. Publisher Full Text\n\nLeimar O, Tullberg B, James M: Mimicry, saltational evolution, and the crossing of fitness valleys.Chapter 16 in: E.I. Svensson and R. Calsbeek, eds. The Adaptive Landscape in Evolutionary Biology,2012;257–270. Publisher Full Text\n\nSavageau MA: Demand theory of gene regulation. II. Quantitative application to the lactose and maltose operons of Escherichia coli. Genetics. 1998;149(4):1677–1691. PubMed Abstract \| Free Full Text\n\nCook SA: The complexity of theorem-proving procedures. In Proc third Annu ACM Symp Theory Comput. – STOC ’71. New York, USA, ACM Press.1971;151–158. Publisher Full Text\n\nLevin LA: Universal enumeration problems (Russian). Probl Peredai Inf. 1973;9(3):115–116. Reference Source\n\nFriedgut E: Sharp thresholds of graph properties, and the k-sat problem. J Am Math Soc. 1999;12(04):1017–1055. Publisher Full Text\n\nMoore C, Mertens S: The Nature of Computation. Oxford University Press,2011. Publisher Full Text\n\nMertens S, Mézard M, Zecchina R: Threshold values of random k-sat from the cavity method. Random Struct Algor. 2006;28(3):340–373. Publisher Full Text\n\nAchlioptas D, Coja-Oghlan A: Algorithmic barriers from phase transitions. In Proceedings 49th FOCS.2008;793–802. Publisher Full Text\n\nGravner J, Pitman D, Gavrilets S: Percolation on fitness landscapes: effects of correlation, phenotype, and incompatibilities. J Theor Biol. 2007;248(4):627–645. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWagner GP, Zhang J: The pleiotropic structure of the genotype-phenotype map: the evolvability of complex organisms. Nat Rev Genet. 2011;12(3):204–213. PubMed Abstract \| Publisher Full Text\n\nRutherford SL, Lindquist S: Hsp90 as a capacitor for morphological evolution. Nature. 1998;396(6709):336–342. PubMed Abstract \| Publisher Full Text\n\nMasel J, Siegal ML: Robustness: mechanisms and consequences. Trends Genet. 2009;25(9):395–403. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJiang P, Kreitman M, Reinitz J: The relationship between robustness and evolution. bioRxiv. 2018;268862. Publisher Full Text\n\nFanti L, Piacentini L, Cappucci U, et al.: Canalization by Selection of de Novo Induced Mutations. Genetics. 2017;206(4):1995–2006. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKlosin A, Casas E, Hidalgo-Carcedo C, et al.: Transgenerational transmission of environmental information in C. elegans. Science. 2017;356(6335):320–323. PubMed Abstract \| Publisher Full Text\n\nGoldschmidt R: The Material Basis of Evolution. Yale Univ.Press, New Haven CT,1940. Reference Source" }	[ { "id": "46602", "date": "23 Apr 2019", "name": "Eors Szathmary", "expertise": [ "Reviewer Expertise Theoretical evolutionary biology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a new paper in a series of exciting papers about phenotypic evolution, based on models of genes genetic regulatory networks, and phenotypes. There are many technical details in the paper that most biologists will find difficult to follow – I would put at least some of these items into an Appendix. Also, I am wondering how many (even theoretical) biologists would readily grasp the meaning of the K-SAT problem as defined. A more accessible formulation (in addition to the strictly technical one) could do a lot for easier understanding (one of the papers by the authors contains a didactive figure, for example). I consider this paper a serious attempt at broadening our views of the role of learning dynamics in evolution, but I think the message could be made a lot clearer. The punch line in a way is in the Discussion that I find both intriguing and baffling. First, as it is portrayed in this paper, the effect of genes is filtered through a genetic regulatory network. It seems that, just like in Ref 401, that this filter itself does not evolve, or at least the dynamics is not given. In other words, how do the weights in the network evolve? Note that in the Watson-Wagner paper in Evolution2 it was the evolution of the weights that was related to Hebbian dynamics. So, how do we stand on this? Second, in this paper the genetic control of the thresholds in also kept implicit, but the evolution of threshold values does play an important role in the Goldschmidtian argument. Is the change in thresholds genetic or epigenetic? If the latter, then it ought to be part of the regulatory network. I guess this is what the statement “gene regulation detects that increase to change the threshold according to simple rules”. How these rules would be implemented in mechanistic terms remains obscure. And how would they arise in evolution? The simple rule implies feedback from expression levels to threshold values. Third, do I understand it correctly that here, in contrast to Ref 401, thresholds can be negative? But a large negative threshold value would mean that expression levels will increase, since negative times negative is positive (Eq. 2.4). If the statement “take three values: a large negative one, a neutral value close to zero and a large positive one” instead refers to gene expression levels rather than threshold values (the latter would in this case be nonnegative) then I do not understand the thought example. Increasing the threshold would reduce the expression level that was selected to go up in the first place! These points seem to me crucial, so their clarification is badly needed. I would appreciate a few numerical examples along with at least hints to answers to the more conceptual questions on the dynamics. Also, according to the cited Nagylaki dynamics3 the population evolves almost as if it were in linkage equilibrium – which cannot hold for the asexual population considered by the authors. There are also some minor issues in the paper: in Eq. 2.4 “s” should be replaced by a sigma, and there are also examples where plurals and singulars in the same sentence do not match.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? No source data required\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "48858", "date": "13 Jun 2019", "name": "Yehonatan Sella", "expertise": [ "Reviewer Expertise Theoretical/Mathematical Systems and Evolutionary Biology." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn the current work the authors study a population genetics model in which fitness is a linear function of a set of phenotypic traits, and where the genotype-to-phenotype map is given by a linear transformation composed with sigmoidal functions. Despite the seeming simplicity of the fitness function, the authors make the case that optimizing fitness is a hard, NP-complete, problem. Under this model, they study the extent to which the fitness landscape (that is, the question of which phenotypic traits contribute positively to fitness, and which contribute negatively) can be inferred from the distribution of these traits in the population after being subject to evolution for a moderate span of time. They then connect this question with that of whether learning algorithms (potentially epigenetic in nature) can help optimize and speed up evolution by learning the fitness landscape.\nToward the goal of inferring the fitness landscape, the authors prove two theorems. Theorem 3.1 concerns a simplified model of an infinite population with no mutations, while Theorem 3.2 concerns a more complex model of a finite population with mutation, stochastic number of descendants and a culling process. In either case, the result is that under certain assumptions, if a mutant genotype is present in the population with high enough frequency after a long enough period of elapsed time, then we can confidently infer that any phenotypic trait differential between the wildtype and the mutant is associated with a higher fitness.\nWhile Theorem 3.1 ignores mutation, even Theorem 3.2 seems at odds with mutation-selection balance. Even a mutant with lower fitness will be present in the population at a frequency that is on the order of the rate of mutation, while the theorem seems to claim that the frequency of such a mutant will be exponentially small with high probability. The authors should resolve this apparent discrepancy. In addition, the main ideas of the proof would be more clearly communicated if the authors would include a treatment of the intermediate model of an infinite population with mutation.\nAs for the discussion of the way learning can speed up the evolutionary process, this part of the paper remains unclear and underdeveloped. The authors discuss a two-step evolutionary process in which the first step consists of small mutations in order to explore the fitness landscape, and the next step involves changing the thresholds involved in the genotype-to-phenotype map in a way that promotes phenotypic traits associated with higher fitness. While this idea is interesting and worth exploring, a few issues arise. A conceptual issue remaining to be addressed is whether the threshold hi is part of the genotype and what mechanism are needed to alter its value. According to the authors, the hi’s can be modified genetically or epigenetically. If epigenetically, it is not apparent what are the environmental cues that will lead to such learning let alone the actual mechanism of modifying them. If genetically, it is similarly unclear in what way the learning of the fitness landscape is being stored, if at all, in the genotype, and what is the connection to the Theorems of chapter 3. The theorems in chapter 3 rely on observing the frequency of a genotype in the population, but such information is not stored in individual genotypes. Additionally, if the thresholds are understood to be variable and subject to selection, then the fitness-maximization problem in fact becomes easy (in contrast to the prior analysis of it as NP-complete), unless we impose restrictions on the range of the threshold. Indeed, one can set the thresholds at positive or negative infinity depending on whether the corresponding trait is positive or negative in order to effectively keep the trait on or off regardless of genotype.\n\nFinally, as this manuscript addresses the relation between learning and the rate of evolution, it would benefit from including a reference to one of the most relevant and intuitive articles written in the late 80’s by Geoffrey E. Hinton & Steven J. Nowlan, “How Learning Can Guide Evolution” Complex Systems, 1, 495-5021. In it, Hinton and Nowlan showed how learning alters the shape of the fitness landscape and thereby provides easier evolutionary paths towards sets of co-adapted alleles. Hinton and Nowlan demonstrated that this effect allows learning individuals to evolve faster than non-learners. Though the learning model presented by Hinton and Nowlan operates at “somatic” timescale, the analogy to mutations at the evolutionary timescale can be drawn.\n\nTo conclude, though it can be improved by the above suggestions, this article touches upon very interesting and important issues in the field of evolutionary biology which are still only lightly investigated, and highlight what might be a fruitful path towards better understanding.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? No source data required\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-358
https://f1000research.com/articles/8-1640/v1	13 Sep 19	{ "type": "Research Article", "title": "Costs and consequences of the Family Nurse Partnership (FNP) programme in England: evidence from the Building Blocks trial", "authors": [ "Kerry Bell", "Belen Corbacho", "Sarah Ronaldson", "Gerry Richardson", "Kerry Hood", "Julia Sanders", "Michael Robling", "David J. Torgerson", "Building Blocks Trial Group", "Belen Corbacho", "Sarah Ronaldson", "Gerry Richardson", "Kerry Hood", "Julia Sanders", "Michael Robling", "David J. Torgerson" ], "abstract": "Background: The Family Nurse Partnership (FNP) is a licensed intensive home visiting intervention programme delivered to teenage mothers which was originally introduced in England in 2006 by the Department of Health and is now provided through local commissioning of public health services and supported by a national unit led by a consortium of partners. The Building Blocks (BB) trial aimed to explore the effectiveness and cost-effectiveness of this programme. This paper reports the results of an economic evaluation of the Building Blocks randomised controlled trial (RCT) based on a cost-consequence approach. Methods: A large sample of 1618 families was followed-up at various intervals during pregnancy and for two years after birth. A cost-consequence approach was taken to appraise the full range of costs arising from the intervention including both health and social measures of cost alongside the consequences of the trial, specifically, the primary outcomes. Results: A large number of potential factors were identified that are likely to attract additional costs beyond the implementation costs of the intervention including both health and non-health outcomes. Conclusion: Given the extensive costs and only small beneficial consequences observed within the two year follow-up period, the cost-consequence model suggests that the FNP intervention is unlikely to be worth the substantial costs and policy makers may wish to consider other options for investment. Trial registration: ISRCTN23019866 (20/04/2009)", "keywords": [ "Randomised controlled trial", "Cost-consequence analysis", "Pregnancy in adolescence", "Prenatal care", "Maternal health", "Home visiting" ], "content": "Introduction\n\nThe Building Blocks trial evaluated the Family Nurse Partnership (FNP) programme, an intensive, nurse-led home visiting programme for young, first time parents who live in areas with a low socio-economic profile. This programme was developed in the USA and was introduced in England by the Department of Health in 2006 with the aim of improving outcomes for health, wellbeing and social circumstances of young first-time mothers and their children. In October 2015 the FNP was transferred from NHS England to Local Authorities (LAs) and it is now provided in approximately 125 different LAs in England. The FNP programme was introduced to be an integral part of the progressive universalism approach recommended in The Healthy Child Programme (HCP). The HCP is delivered by the Family Nurse rather than by health visitors for women who enrol onto the programme.\n\nDespite an extensive evidence base in the US, very little was known about the generalisability to a UK context which implements a very different health care model to that in the US. Hence, the Building Blocks (BB) trial was commissioned to evaluate the effectiveness and cost-effectiveness of the FNP intervention when delivered in a comprehensive publicly funded health care setting. The results from the effectiveness analysis showed no statistically or clinically significant difference associated with FNP for any of the four primary outcomes: smoking cessation (adjusted OR 0·90, 97·5% CI 0·64–1·2), birth weight (adjusted mean difference 20·75 g, 97·5% CI –47·73 to 89·23), second pregnancies within two years (AOR 1·01, 0·77–1·33), or child A&E attendances and admissions to hospital (AOR 1·32, 97·5% CI 0·99–1·76, p=0.03)1.\n\nNICE has long recommended the use of cost-utility analysis as the primary means of assessing cost-effectiveness where health is the sole or dominant benefit of influence. However, with the move to local authority responsibility in 2013, NICE broadened its approach to the appraisal of public health interventions to place more emphasis on cost-consequence and cost-benefit analyses to supplement the cost-utility analysis for complex interventions2. Unlike the cost-utility analysis which reports only on the number and cost of QALYs gained, the cost-consequences analysis (CCA) is able to present disaggregated costs and a range of outcomes which allows readers to form their own conclusions on relevance and relative importance to their own decision making context3. This is particularly valuable for evaluation interventions with an array of health and non-health benefits that cannot be measured in a common unit. CCAs are not restricted to a particular viewpoint hence decision makers can estimate the impact of their decisions on sectors beyond health, such as education and criminal justice4. Furthermore, CCA is able to take many items into account that local authorities are likely to find important, including the trade-off between long-term goals and a paucity of short-run funding, and spill over effects into other areas of local government responsibility.\n\nGiven the complexity of the FNP programme and the diverse range of policy relevant outcome measures, a multi-faceted approach was taken to evaluate the programme’s cost-effectiveness. A cost-utility analysis conducted alongside the trial showed only nominal benefits in terms of maternal QALY gains which were not considered cost-effective based on NICE (National Institute for Health and Care Excellence) acceptability thresholds5.\n\nThe analysis aims to list all relevant health and non-health related resource use, largely public or third sector providers of universal and specialist services locally accessible to teenagers of young children, and the costs associated with each of the trial arms as well as the consequences of the trial, specifically, the primary outcomes. In the UK, health care policy decisions are largely driven by the National Institute for Health and Care Excellence (NICE) which base their recommendations on the number of quality adjusted life years (QALYs) that can be gained by an intervention. As the children in the participating families were too young to draw QALY estimates, the overarching perspective of the wider economic analysis that was conducted alongside the trial could only be completed from the perspective of the mother due to availability of data. In keeping with this approach, the cost-consequence analysis also considered only resource use of the mother.\n\n\nMethods\n\nThe Building Blocks trial was a two-arm pragmatic, non-blinded, parallel-group, randomised controlled trial (RCT) which recruited within a community midwifery setting at 18 partnerships between LAs and primary and secondary care organisations in England. A total of 1645 participants were randomised in the study. The present analysis is based on the 1618 participants recruited to the trial that did not later become ineligible (e.g. due to miscarriage) or withdraw their consent. Participants in the Intervention arm (n=808) received home visits from the FNP nurse during their pregnancy and in the two years following childbirth. Participants allocated to the usual care arm (n=810) received care from the local maternity and health visiting services in line with usual practice. Details on recruitment and follow-up rates are described in full elsewhere1. Informed consent was obtained from all individual participants included in the study.\n\nThe primary outcome measures for the trial were tobacco use at late pregnancy (34–36 weeks’ gestation), birth weight, emergency attendances and hospital admissions for the infant within 24 months of birth, and the proportion of women with a second pregnancy within 24 months post-partum. There were also a large number of secondary outcomes across a number of domains; socio-economic, maternal health and well-being, health behaviours, pregnancy and birth, social support and the use of services. For all primary outcomes a 97.5% confidence interval (CI) and p-value was presented; for all secondary outcomes a 95% CI and p-value was presented. The outcome measures are described in full elsewhere1.\n\nData for outcomes and resource use were collected by self-reported questionnaires at various time-points throughout the trial, specifically; baseline, late pregnancy (34–36 weeks gestation) and 6, 12, 18, and 24 months postpartum. Baseline and 24 month data were collected by face-to-face interview by a locally based researcher whilst data at other time points were collected via telephone by qualified telephone interviewers. Additionally, data related to the use of health care services for each trial participant were collected from Hospital Episode Statistics (HES) via NHS Digital and primary care (general practitioner (GP) records). Birth data were collected by records abstraction from maternity records.\n\nThe mean numbers of the health and non-health care items per woman are presented for both groups for the duration of the trial in their respective units, e.g. mean number of GP visits, average number of weeks in education etc. The difference between the two groups for these items is also presented. The totals for each item (e.g. total number of GP visits) have been calculated for each participant in both groups over the duration of the trial (i.e. from baseline up to the 2-year interview).\n\nResource use is averaged across all women in each arm leading to small mean numbers for resource use. Median, minimum and maximum values are reported to highlight the highly skewed nature of resource use. Costs were largely associated with only a minority of women using each resource. Resource use and associated costs are reported in terms of the mother only. All consequences of the intervention, for both mother and child, are reported.\n\nUnit costs applied to the resource use were retrieved from several sources. Unit costs pertaining to health care resource use (Table 1) were sourced from NHS reference costs6 and Personal Social Service Research Unit (PSSRU) Unit Costs of Health and Social Care 20137. However, as no unified document currently exists to provide unit costs for the non-health related items, these were sourced from other relevant sources such as governmental websites, research documents and reports. Given that costs were taken from a range of sources the years of pricing vary for each item. These are reported in Table 2. No attempt was made to discount/ inflate these items to a consistent rate as appropriate inflation rates for this type of resource are not available. It is thus acknowledged that current figures per resource use may differ slightly from those documented.\n\n^ PSSRU unit costs are based on mean full-time basic salary; including overheads and qualifications.\n\nFNP - Family Nurse Partnership, PSSRU - Personal Social Service Research Unit\n\nFor the purpose of the cost-consequence analysis, multiple imputation analysis was considered unnecessary as the model serves only to highlight likely resource use and the average costs per resource. This decision was further vindicated by the small number of women who reported using each of the resources, which in some cases were less than five women. Such small numbers of affirmative responses would prevent imputations from running effectively and result in high error rates. Instead, several assumptions were made to reduce the amount of missing data. Where a particular answer to a question meant that sub-questions were skipped, a value of zero was attributed to the sub-questions. For example, respondents were first asked a yes or no question as to whether they had utilised a particular resource, then how often following an affirmative response. Where respondents stated that they had not used a resource, frequency of use was assumed to be zero. Conversely, where a positive response was noted but not frequency value was given, a value of one was assumed as a conservative estimate. Due to the high number of variables and low number of complete cases, it was not considered pragmatic to summarise each variable by complete case and adjusted values.\n\nThe mean resource use was calculated for each resource use item according to the appropriate unit (e.g. mean number of days). The incremental difference between trial arms was calculated for each item of resource use by subtracting the values of participants allocated to receive usual care from the values of participants who were allocated to receive the intervention (FNP). Mean unit costs were calculated by item for each of the trial arms by multiplying the resource use by an appropriate unit costs (see Table 3 for unit costs). Resource use costs were summed by type, e.g. health related resource use or non-health related resource use. All analyses were conducted in Stata version 128. The consequences presented in Table 3 are the primary outcomes of the Building Blocks trial and full details of how these were calculated are presented in primary trial paper9.\n\nFNP - Family Nurse Partnership\n\n\nResults\n\nUnit costs of all health and non-health related resources used by the mother only are presented in Table 1 and Table 2. The results of the cost-consequences analysis can be seen in Table 3. For each cost item, the mean resource use per participant for both the FNP arm and the usual care arm are displayed, alongside the incremental resource use. The mean cost per participant is also shown for each item according to trial arm, as well as the incremental cost. In terms of the consequences, the primary outcomes of the Building Blocks trial are listed for both the FNP arm, usual care arm and the difference between the two. A descriptive summary of the costs and consequences is provided, whereby items are discussed in a disaggregated format for the broad range of costs and consequences (spanning health care and non-health care) included in the analysis.\n\nLittle difference was observed in terms of average health care resource use across the majority of services though consistently small differences were found between trial arms in favour of FNP. The largest differences in terms of resource use were the average number of A&E attendances and the mean number of health visitor visits received, both of which, as anticipated, were lower for women receiving the FNP intervention compared to women under usual care.\n\nAs only small differences in average resource use were observed, the average per woman spend on each resource reflected only slightly notable differences. Overall, average total health-related resource use costs were slightly lower for the FNP arm compared to usual care. Lower average costs were found in the FNP arm in relation to inpatient care, outpatient attendances not related to maternity services, A&E attendances, and midwife and health visitor visitations. The largest difference in costs was found in relation to inpatient care requiring one or more nights in hospital. In this instance, the mean difference between arms was £307 with lower costs being associated with FNP recipients. A difference was also found between the average cost per participant in terms of community-based maternity and early years services, specifically health visitor and midwife visitation. On average £82 less was spent per woman receiving the FNP intervention compared to usual care for health visitor visits (both home and clinic based) whilst £21 less was spent per woman on midwife visits (both home and clinic based) reflecting a mean reduction in spend of just over £100 per woman on these key early years services. The number of health visitor visits received by women was lower in the FNP arm than the usual care arm as visits from FNP nurses were received in place of standard health visitor visits. Consequently, though the cost attributed to health visitor visits appears reduced, there is the additional cost of the FNP nurses. The cost of the health visitors provided in usual care would thus not be saved through the FNP intervention but would be spent instead on specific FNP trained nurses. Only minor differences were noted in terms of cost associated with inpatient day admittances not requiring an over-night stay (-£6.51), non-maternity based outpatient attendances (-£4.86), and A&E attendances (-£5.73), with all being in favour of FNP. Omitting the intervention costs, a total of £8, 872 an average was spent on health care for women receiving the FNP intervention compared to a total average of £9, 162 per woman in the usual care arm, a difference of £290 per woman between the two arms.\n\nIn addition to health care resource use, there was very little difference in the non-health care resource use between the two groups. The average cost of non-health related resource use for women receiving the FNP intervention was £1,866 compared to £1,738 for women under usual care, an average difference of only £128 per woman in favour of the control group. In general, reported uptake of all non-health related resources was fairly low in both groups as can be seen through the mean and median values presented in Table 3. For the majority of resource items, the median resource use for both groups is zero, confirming a low uptake. Mean values are thus influenced by only a minority of women.\n\nParticipation in education and training was also similar between groups for both mainstream schooling and dedicated support units with little difference in resource use or mean costs.\n\nIn terms of supportive services, children’s centres, such as Sure Start centres, were the most widely used service by both groups, though this averaged at only 2 visits per woman in each of the trial arms. Mean costs of children’s centre use was slightly higher in the FNP arm. This is likely a consequence of outlying values as the maximum number of visits was considerably higher in the FNP arm compared to usual care (38 and 29).\n\nResource use of childcare was found to be particularly low at all time-points and overall mean resource use per woman, within both groups, was particularly low. Slightly higher costs were found in FNP women in line with the marginally higher rates of resource use. The most widely used type of childcare was crèche/day nursery within a school or college though the mean number of weeks used throughout the follow-up period was very low at only 1.35 weeks for FNP women and 1.10 weeks for usual care women. As this resource use was low the total resource use was low resulting in a low average cost per woman of £5.79 per FNP woman and £4.71 per usual care woman though the cost for women actually using childcare would be much higher.\n\nResource use for foster care and temporary accommodation was also low in terms of the number of women reporting usage of these resources for both groups of women. Average costs are thus driven by a minority of women. Average costs for temporary accommodation were higher for FNP women than usual care women for teenage parent accommodation, supported accommodation and mother and baby units due to increased use of these by FNP women. The largest average difference between groups was for teenage parent accommodation, with FNP women costing on average £75.73 more than usual care women; again this was due to higher levels of resource use in the FNP group.\n\nPrimary outcomes. No difference was observed in the rate of smoking in late pregnancy between the two arms (adjusted OR: 0.90, 97.5% CI: 0.64 to 1.28). There was no difference in the reported number of cigarettes smoked during late pregnancy for participants classified at baseline as smokers (adjusted difference in means, FNP-Usual Care: 0.119 cigarettes, 97.5% CI: -0.73 to 0.97). On average, birth weights were marginally higher in the intervention arm compared to the control arm though this was not statistically significant (adjusted difference in means, FNP-Usual Care: 20.75 grams, 97.5% CI: -47.73 to 89.23). There was no difference in the proportion of women reporting a second pregnancy within the two years of their first child’s birth (adjusted OR 1.01, 97.5% CI: 0.77 to 1.33). Rates of child emergency attendances and admissions for any reason in secondary care prior to their second birthday were higher in the intervention arm though the difference was not statistically significant (adjusted OR 1.32, 97.5% CI: 0.99 to 1.76). There were no differential effects due to age, deprivation, participation in employment, education or training, or basic life skills for any of the primary comparisons.\n\nSecondary outcomes. Some benefit of the intervention was observed in terms of maternally rated language developmental delay in the children at both 12 (OR 0.50, CI: 0.35 to 0.72) and 18 months (OR 0.66, CI: 0.48 to 0.90). This was further reflected in the Early Language Milestone assessment at the end of the trial period which showed higher scores on average for children in the intervention arm (adjusted difference of means, 4.49, CI: 0.52 to 8.45) indicating more advanced levels of language development. Benefits of the intervention were also observed for breastfeeding intentions (adjusted OR 1.32, CI: 1.02 to 1.70) though this did not translate into a difference in the actual proportions of mothers initiating breastfeeding between the two arms.\n\nA greater proportion of children in the intervention arm attended an Emergency Department for an injury or ingestion at 6 months (adjusted OR 1.52, CI 0.86 to 2.70) and 12 months (adjusted OR 1.16, CI 0.92 to 1.46), were referred to social services by two years (adjusted OR 1.27, CI: 0.93 to 1.73) and had a safe guarding event recorded in their GP record (adjusted OR 1.85, CI: 1.02 to 2.85) compared to children in the control group. These differences were not statistically significant.\n\nNo evidence of a difference was observed for any other maternal or parenting and child outcomes assessed in the trial.\n\nAt two years, mothers in the intervention arm reported lower rates of not being in employment, education or training (NEET) than the control arm, though this was not statistically significant. A larger proportion of participants in the intervention arms reported a maximum level of social support at 18 months postpartum (25.7%) compared to those in the control arm (20.3%) with a similar difference being maintained at 24 months. A small difference between the arms was observed over the whole follow-up period (adjusted OR 1.5, CI: 1.06 to 2.12). Fewer participants in the intervention arm reported ever being homeless during the study period compared to the control arm (adjusted OR 0.76, CI: 0.55 to 1.05) and higher levels of self-efficacy were observed in the intervention arm (adjusted mean difference 0.44, CI: 0.10 to 0.78).\n\nNo evidence of a difference was observed for any other parental life course outcomes assessed in the trial.\n\n\nDiscussion\n\nThis analysis shows limited differences between the resource uptake and associated costs between the two trial arms (FNP intervention and usual care). As demonstrated in Table 2, for most of the health care resource use items, average costs were slightly lower for the FNP arm compared to usual care. Conversely resource use and associated average per person costs were higher for non-health related items. This however is not necessarily a negative outcome as some of the items associated with greater costs were positive aspects such as attendance at a mainstream education programme and attendance at children’s centres. Despite these resources being considered positive outcomes it is important to remember that these are still additional costs which could be incurred as a result of the FNP intervention and should be considered in policy making. It is also worth noting that although there were more child A&E attendances and admissions associated with families receiving the intervention, this is unlikely to be a detrimental effect of the intervention. Instead this probably reflects an increased awareness of mothers receiving the intervention, and their increased contact with a health professional could mean that health concerns are identified more readily causing mothers to seek more help.\n\nPrevious studies of the FNP programme in the United States (US) have shown positive results in terms of improved prenatal health, childhood injuries, subsequent pregnancies, maternal employment and child school readiness10–12. The Building Blocks study is the first randomised controlled trial of FNP in the UK. The statistically insignificant effects observed in the present work which are discussed in the effectiveness paper9 likely reflect the differences in health care provision between the US and the UK. In the US, low income mothers are not able to access many health care and social services. In contrast, all pregnant women in the UK can access a wide provision of maternal care including community care family doctors, midwives and specialist public health nurses. It is thus likely that the FNP programme does not offer a sufficiently enhanced service to the level of usual care provided in the UK. This study thus adds to the knowledge base of FNP and is useful to policy makers considering the value of implementing FNP within a UK-style health care system.\n\nAs a methodological approach, cost-consequence analysis is generally limited by its dependence on subjective decision making by policy makers, as it relies on individual judgement of what consequences are sufficiently beneficial to merit investment. The lack of a comparative statistical component could also be viewed as a limitation. However, the cost-consequence model is of value to provide a simplistic overview of costs associated with the intervention, particularly those that are not obvious such as youth offending services and family information services. A cost-consequence balance sheet was implemented to provide policy makers with a clear descriptive summary of the health and non-health related costs associated with the FNP intervention. This offers an advantage over more technical methods of economic analysis. Furthermore, though the use of QALYs is generally favoured by clinical decision making bodies such as the National Institute for Health and Care Excellence (NICE)13, the calculation of QALYs may not always be sensitive in particular populations. When QALY gains were assessed in a cost-utility analysis conducted as part of the BB trial, no significant gains were identified and the intervention was deemed not cost-effective when considered from the perspective of the mother14. Given the dimensions of the EQ-5D (pain, mobility, self-care, usual activities, and depression/anxiety) it may be the case that the instrument was insensitive to the population and the intentions of the intervention to primarily improve child health outcomes rather than maternal health outcomes thus giving further weight to the case for considering outcomes beyond health.\n\nThe analysis considered only resource use of the mother; this is due to the overarching perspective of the wider economic analysis that was conducted alongside the trial, the primary analysis being a cost-utility analysis. Currently there is no accepted instrument for measuring utility in children as young as those in the Building Blocks trial hence the cost-utility analysis could only take the perspective of the mother. Future research could incorporate child resource use but given the absence of statistically significant results for any of the Building Blocks trial primary outcomes, this would not change the overall message of the work and hence was deemed unnecessary.\n\nAs an approach, cost-consequence analysis cannot usually conclude whether an intervention is worth funding as the judgement usually lies upon the gains that would likely be attained by recipients. However, given that the FNP intervention does not bring about any significant improvements to recipients compared to usual care on the basis of the short-term outcomes, together with the higher cost of delivering the intervention compared to usual care (£1812 more per woman), these two factors suggest that based on current evidence the continuation of the programme cannot be justified.\n\nIn all, the cost-consequence model summarises the between-group differences in resource use and costs, alongside the primary trial outcomes. As can be seen, there are a large number of potential factors that are likely to attract additional costs beyond the obvious implementation costs of the intervention. Given these costs and the way in which only small benefits were observed in the trial analysis when considering primary outcomes alone, the cost-consequence model provides further support that the FNP intervention is unlikely to be worth the substantial costs and policy makers may wish to consider other options for investment.\n\n\nEthical approval\n\nThe trial is registered with International Standard Randomised Control Trial Number ISRCTN23019866 on 20 April 2009. The Building Blocks trial protocol and all amendments were reviewed and approved by the Wales NHS Research Ethics Committee (09/MRE09/08). All procedures were in accordance with the ethical standards set out in the 1964 Helsinki declaration and its later amendments.\n\n\nData availability\n\nThe Building Blocks Trial dataset comprises data from multiple sources including third party. The Hospital Episode Statistics (HES) data, which forms a substantial component of the present analysis, is supplied under licence from NHS Digital. The HES dataset is not currently available for analysis to researchers outside Cardiff (the lead institution) due to the information provided to participants at consent (which identified which institutions were acting as data controller / processors).\n\nResearcher can apply for access to the full dataset by contacting Cardiff Centre for Trials Research (CTR) (see: https://www.cardiff.ac.uk/centre-for-trials-research/about-us/data-requests). The restriction on HES data (originally supplied via NHS Digital) applies. CTR will consider whether section 251 approval would enable release of participant identifiers to apply directly to NHS Digital for the data used in the trial.\n\nThe results of the Building Blocks trial are presented in full at: https://www.cardiff.ac.uk/__data/assets/pdf_file/0009/504729/Building-Blocks-Full-Study-Report.pdf\n\nThis publication includes the full CONSORT statement for the trial on page 564.", "appendix": "Acknowledgments\n\nThe South East Wales Trials Unit (SEWTU) is funded by the Wales Assembly Government through the Health and Care Research Wales and the authors gratefully acknowledge SEWTU’s contribution to study implementation. We thank all the women who participated in the study, the local professionals who facilitated the recruitment and study implementation, and the family nurses who delivered the intervention. We acknowledge all the other contributors to the study who are listed in full study report. We thank all the trial steering committee independent members: Ann Louise Kinmonth (Chair), Silvia van den Heijkant, Pamela Park, Stavros Petrou, Rachel Tonkin; and the data monitoring committee independent members: Gordon Taylor (Chair), Lucy Akhtar, Sara Kenyon. We would like to pay special tribute to the late professor Paul Wainwright who has the initial Chair of the Data Monitoring Committee. We thank the stakeholder involvement work package members: Joyce Kenkre (lead), Lily Bidmead, Kamila Hawthorne, Lesley Lowes, with contributions from members of the Books & Babies Group, and the Young Mums Groups. The trial administrators were Jackie Swain, Katy Addison, and Rhys Thomas.\n\n\nReferences\n\nRobling M, Bekkers MJ, Bell K, et al.: Effectiveness of a nurse-led intensive home-visitation programme for first-time teenage mothers (Building Blocks): a pragmatic randomised controlled trial. Lancet. 2016; 387(10014): 146–155. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNational Institute for Health and Care Excellence: Methods for the Development of NICE Public Health Guidance [Internet]. (third edition), London: NICE. 2012. PubMed Abstract\n\nDrummond M, Sculpher MJ, Torrance GW, et al.: Methods for the economic evaluation of health care programmes. 3rd edition, Oxford, UK: Oxford University Press, 2005. Reference Source\n\nBrazier J, Ratcliffe J, Salomon J, et al.: Measuring and Valuing Health Benefits for Economic Evaluation. Oxford: Oxford University Press. 2007. Reference Source\n\nCorbacho B, Bell K, Stamuli E, et al.: Cost-effectiveness of the Family Nurse Partnership (FNP) programme in England: Evidence from the building blocks trial. J Eval Clin Pract. 2017; 23(6): 1367–1374. PubMed Abstract \| Publisher Full Text\n\nDepartment of Health: NHS Reference Costs 2012/13. 2013. Reference Source\n\nCurtis L: Unit Costs of Health and Social Care 2013. Personal Social Services Research Unit: University of Kent. 2013. Reference Source\n\nStataCorp: Stata Statistical Software: Release 12. StataCorp LP: College Station, TX. 2011. Reference Source\n\nRobling M, Bekkers MJ, Bell K, et al.: Effectiveness of a nurse-led intensive home-visitation programme for first-time teenage mothers (Building Blocks): a pragmatic randomised controlled trial. Lancet. 2016; 387(10014): 146–155. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKitzman H, Olds DL, Henderson CR Jr, et al.: Effect of prenatal and infancy home visitation by nurses on pregnancy outcomes, childhood injuries, and repeated childbearing. A randomized controlled trial. JAMA. 1997; 278(8): 644–52. PubMed Abstract \| Publisher Full Text\n\nKitzman H, Olds DL, Sidora K, et al.: Enduring effects of nurse home visitation on maternal life course: a 3-year follow-up of a randomized trial. JAMA. 2000; 283(15): 1983–9. PubMed Abstract \| Publisher Full Text\n\nOlds DL, Kitzman H, Hanks C, et al.: Effects of nurse home visiting on maternal and child functioning: age-9 follow-up of a randomized trial. Pediatrics. 2007; 120(4): e832–45. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNational Institute for Health and Care Excellence: Guide to the methods of technology appraisal. London: NICE. 2013. Reference Source\n\nCorbacho B, Bell K, Stamuli E, et al.: Cost-effectiveness of the Family Nurse Partnership (FNP) programme in England: Evidence from the building blocks trial. J Eval Clin Pract. 2017; 23(6): 1367–1374. PubMed Abstract \| Publisher Full Text" }	[ { "id": "53944", "date": "01 Oct 2019", "name": "Shaun Harris", "expertise": [ "Reviewer Expertise Health Economics" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors present a cost-consequences analysis of a family nurse (FNP) led intervention (versus usual care) to support teenage mothers in England. The analysis is based on approximately 1600 participants recruited to the Building Blocks RCT. The randomised design and size of the study should provide robust results for policy makers.\nGiven the diverse range of costs and outcomes a cost consequence analysis is a reasonable and appropriate approach to take. The pragmatic approach to address missing data also appeared to be reasonable given the context. Key outcomes and cost-utility data from this trial have already been reported (Robling et al. 20161; Corbacho et al. 20172). Therefore, the contribution of this paper is largely in providing additional details about health (and wider) resource use in each arm of the trial.\nThe results show that there is little difference in either resource uptake or associated costs between FNP and usual care. The authors conclude that the intervention is unlikely to be cost-effective. The authors suggest that previous positive results from US studies might not be replicated in England due to more extensive existing access to maternity care in the UK.\n\nThere are several issues that should be clarified:\nIn Table 3, it would be useful to report the total healthcare costs to establish the aggregate cost differences (and 95% CI) between FNP and usual care. These total costs are briefly mentioned on page 9 (paragraph 2) but do not appear in Table 3.\n\nThe cost of the FNP intervention itself should be included in Table 3. These are only introduced in the discussion section currently. Inclusion in Table 3 should allow the authors to estimate the net incremental healthcare costs of FNP.\n\nAlthough there is no one agreed approach to inflating costs across difference sectors of the economy, the approach taken by the authors (of not standardizing cost years) seems unsatisfactory as it does not allow the estimation of total costs (healthcare and other sectors) that would be useful for readers/policy makers. A GDP deflator index - while not perfect - would be a better starting point.\n\nPage 9 (paragraph 3): median values are not presented in table 3.\n\nA brief description of the intervention theory and aims and the eligibility criteria would help this article 'stand alone' from previous articles.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "53943", "date": "08 Oct 2019", "name": "Jacqueline Barnes", "expertise": [ "Reviewer Expertise Evaluation of programmes to support disadvantaged parents and children", "The relevance of communities for parenting" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe Building Blocks team are to be congratulated for taking a cost-consequences approach (CCA) rather than the cost utility method relying on QALYs, the strategy frequently applied to interventions. The CCA is much more relevant for this type of trial, with many non-medical potential benefits included in the secondary outcomes. It is well known that the long-term impact of early intervention such as the Nurse Family Partnership (NFP, known as FNP in the UK) can be within the areas of education, employment or criminal justice. Unfortunately for the team they currently have information only up to the end of programme delivery when children were 24 months of age, and are reliant of the primary outcomes being medical in nature. Many of the most well documented impacts of the NFP intervention, from trials in the USA, have been identified when the children are in primary or secondary school, or during young adulthood. It is hoped that further follow-up of the trial participants will be feasible. The paper acknowledges this point.\nThe design is strong, albeit limited due to lack of sufficient costing of services used by infants, meaning that all of the services used are those use by the mother. Thus it is not clear whether a visit to a children's centre includes any services explicitly experienced only by the child are covered. One point about the costs applied to Children's Centre visits is that they are based on the economic evaluation of Sure Start Local Programmes (SSLPs). While Children's Centres, where if they still operate, are largely derived from SSLPs, the range of services is likely to be fewer meaning that costs may be in fact lower.\nAll the details of the method are clearly presented, meaning that the work could be reproduced if there was access to the full dataset. This is not possible since participant consent precluded anyone not a member of the Building Blocks team having access to the Hospital Episodes Statistics data. This is clearly stated in the paper.\nThe discussion is very sensible, noting that aspects of the mothers' experiences such as attending an educational establishment or a children's centre, while incurring costs at the time are likely to lead to less of a cost to society in the future if this leads to better employment prospects, or a child more ready to start school. Similarly more women and infants gaining access to housing rather than being homeless is of great benefit to society and to local communities. Thus it is disappointing that the paper concludes that continuation of the programme cannot be justified. This strong conclusion is not really warranted until the children can be followed until the beginning of formal education at the very least. The FNP children had some benefits in terms of language development so it is important to identify whether fewer required the support of speech and language therapists or educational psychologists and special education teachers. Thus it would have been useful to have some provisos included in the conclusion. While it is not possible in relation to the Building Blocks RCT, it would also be useful to comment that further research with a sample of women identified using additional risk indicators beyond age and first pregnancy, and one involving FNP practitioners with substantial experience rather than newly trained Family Nurses, would add to the capacity for commissioners to decide about continuing to support the FNP programme.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-1640
https://f1000research.com/articles/8-1639/v1	13 Sep 19	{ "type": "Study Protocol", "title": "Evaluation and pilot implementation of essential interventions for the management of hypertension and prevention of cardiovascular diseases in primary health care in the Republic of Tajikistan", "authors": [ "Dylan R.J. Collins", "Tiina Laatikainen", "Mekhri Shoismatuloeva", "Isfandiyor Mahmudzoha", "Zakriya Rahimov", "Dilorom Sultonova", "Bunafsha Jonova", "Jill L. Farrington", "Tiina Laatikainen", "Mekhri Shoismatuloeva", "Isfandiyor Mahmudzoha", "Zakriya Rahimov", "Dilorom Sultonova", "Bunafsha Jonova", "Jill L. Farrington" ], "abstract": "Background: Non-communicable diseases (NCDs) are the leading cause of death worldwide and are a major burden in Tajikistan. The health system of Tajikistan is still shaped by the country's Soviet legacy and the pace of reform has been slow, with high patient out-of-pocket expenditure. The aim of this study is to determine the feasibility of implementing and evaluating essential interventions for the management of hypertension and prevention of cardiovascular disease in primary health care in Tajikistan. Methods and analysis: A pragmatic, sequential mixed methods explanatory design, composed of quantitative and qualitative strands will be used with greater weighting of the quantitative strand. A single geographic district was nominated by the Ministry of Health and chosen for implementation. All primary health care centres in the district that meet inclusion criteria will be included; half will be randomly assigned to the intervention arm and half to the control arm. The overall process is organized into seven steps: (1) refresh clinical decision-making tools including open source WHO PEN and HEARTS resources; (2) update training package for primary health care workers; (3) collection of baseline data; (4) training staff in intervention clinics; (5) implementation of protocols and implementation coaching; (6) collection of follow-up data after 12 months; (7) evaluation of results and sharing experience. Ethics and dissemination: Ethical review and approval have been obtained. Findings will be disseminated at the participant level, national level through a national conference of key stakeholders, and internationally through publication in an open-access peer review journal.", "keywords": [ "Cardiovascular disease", "Tajikistan", "Hypertension", "Public Health" ], "content": "Introduction\n\nNon-communicable diseases (NCDs) are the leading cause of death worldwide, causing 41 million deaths in 20161. Half were due to cardiovascular diseases (CVD)2. Primary health care (PHC) systems are an essential component required to tackle this global burden3. While strong PHC systems based on the principles of family medicine contribute to achieving universal health coverage, better health outcomes and economic and social development, many nations lack PHC capacity4,5. To help national governments develop primary health care capacity for NCDs, the World Health Assembly endorsed the WHO Global Action Plan for the Prevention and Control of NCDs 2013–20206,7.\n\nA country of the former Soviet Union, Tajikistan is located in Central Asia; the capital city is Dushanbe. Ranking among the poorest in the WHO European Region, Tajikistan had a gross domestic product (GDP) of 796 USD per capita in 20168. The WHO Regional Office for Europe continues to support Tajikistan to implement the Global Action Plan, including support for the development, testing, and national scale-up of evidence-based clinical guidance and health policy for the prevention and management of NCDs.\n\nNCDs are a major burden in Tajikistan. The age-standardized NCD mortality rate was 685.3 per 100,000 in 20159. In 2016, the probability of dying prematurely (between the ages of 30 and 70 years) from CVD, cancer, diabetes, or chronic respiratory disease was 25.3%; like most populations, this rate is higher for men than women10.\n\nThe Ministry of Health of Tajikistan and the World Health Organization Regional Office for Europe recently completed a population STEPS survey (unpublished by WHO) which indicated that 25.7% of men are current users of tobacco (including smokeless tobacco) although rates are low amongst women (0.2%)11. For cultural reasons, alcohol use is relatively low with only 9.4% of men and 0.2% of women identified as current drinkers, and lifetime abstention is very high. The prevalence of obesity is 11.9% in women and 15.4% in men, and around half (46.7%) the adult population are overweight.\n\nThe prevalence of raised blood pressure (BP) (defined as systolic BP ≥ 140 mmHg and/or diastolic BP ≥ 90 mmHg or currently taking medication for raised blood pressure) among the adult population is 32.2%. Less than half of men (44.3%) have never had their blood pressure measured; the same is true for 24.7% of women. Nearly all (96.9%) adults have never had cholesterol measured. One-in-seven (13.8%) people aged 40–69 years have a 10-year fatal or non-fatal CVD risk of over 30% (including those with an existing CVD).\n\nThe health system of Tajikistan is still shaped by the country's Soviet legacy and the pace of reform has been slow12. The country has the one of the lowest total health expenditure (THE) per capita in the WHO European Region, estimated to be 6.88% of GDP in 2014, and the proportion of THE that is financed privately through out-of-pocket payments (OOP) is 61.69%; second highest in the WHO European Region8. Although formally free, OOP expenditure for PHC is common and expenditure on pharmaceuticals is the biggest financial burden.\n\nRecent reforms have aimed to strengthen PHC, but success has been mixed. In 2010, Tajikistan launched the National Health Strategy for the period 2010–202013, which recognized the importance of health system strengthening, and highlighted the development of PHC based on family medicine practice as a top priority. The National Programme on the Development of Family Medicine 2011–2015 in Tajikistan had the goal of ensuring the sustainable development of PHC according to the principles of family medicine. Some substantial improvements were achieved, including trainings for the health workforce, the review of clinical protocols, promotion of quality of care, increasing capacity in family medicine, improvements in evidence-based practice, and greater availability of resources. Nevertheless, challenges have been highlighted, including insufficient pace and scale of initiatives, and the scope of work and distribution of family doctors across the country14. Although some computers have been installed in district health care facilities to introduce electronic submission of statistical reports, the vast majority of rural health clinics rely on paper records which need to be completed manually, hindering reliable data collection, processing and analysis12.\n\nIn Tajikistan, health services are delivered at five levels: rural (village), district (rayon), city, oblast (regional), and republican (national). In villages, PHC services are provided in rural health centres (RHCs) with a family medicine doctor or health houses with feldshers, family medicine nurses and midwives.\n\nIn order to build capacity in PHC and ultimately prevent premature mortality from major NCDs, from 2013 to 2015 Tajikistan endeavoured to adapt and pilot the World Health Organization Package of Essential NCD Intervention for Primary Healthcare in Low Resource Settings (WHO PEN)5. WHO PEN includes simplified clinical protocols which together cover the integrated management of hypertension and diabetes, as well as education and counselling on healthy behaviours aimed to prevent CVD. The central strategy of this integrated approach is the use of total cardiovascular risk assessment to stratify and target individuals at high CVD risk, a process considered to be a “best buy” intervention by the WHO15.\n\nThese early efforts led to limited success. A national PEN steering group was established in 2012, facilities were assessed for access to essential medicines and technologies, national clinical guidelines were updated to include the WHO/ISH risk prediction chart16, and family doctors and nurses from 9 pilot facilities received two-day training on the WHO PEN protocols during 2013/14. Apart from a “refresh” training for selected clinics two years later, there was no further intervention, and the lack of systematic monitoring and evaluation made it difficult to demonstrate any significant change in clinical practice, disease detection, disease control rates or clinical outcomes.\n\nIn 2016, Tajikistan agreed to embark on a pilot of the global WHO HEARTS initiative supported by WHO17. Work on this was able to proceed in earnest in 2018 with the publication of the HEARTS technical package which aimed to provide a strategic approach to improving cardiovascular health in countries18. It comprises six modules and an implementation guide to support Ministries of Health to strengthen CVD management in primary health care settings.\n\nLed by the Ministry of Health and Social Protection, the WHO PEN steering group was expanded to include chief specialists, district and rural health facility managers and clinicians, the Republican Clinical and Training Center of Family Medicine and the Service of State Supervision for Medical Activities and Social Protection of the Population of the Republic of Tajikistan. The national steering group is supported by an international team of experts coordinated jointly by the WHO Regional Office for Europe and the WHO Country Office in Tajikistan. Tajikistan has a particular interest in systems monitoring, and in developing sustainable and scalable approaches, building upon other initiatives in PHC reforms and quality improvement.\n\n\nAim and objectives\n\nThe aim is to determine the feasibility of implementing and evaluating essential interventions for the management of hypertension and prevention of CVD in PHC in Tajikistan.\n\nIn order to achieve this aim, the objectives are to:\n\n1. Determine the baseline performance of PHC services with respect to essential interventions for the management of hypertension and prevention of CVD;\n\n2. Assess the ability to conduct practical trainings for primary care doctors and nurses and implement Tajikistan-adapted HEARTS tools, in one pilot region of Tajikistan;\n\n3. Estimate the change in clinical practice with respect to essential interventions for the management of hypertension and prevention of CVD after 12 months of implementation;\n\n4. Determine the feasibility of collecting quantitative data required for future studies of effectiveness from routine clinical data.\n\n\nMethods and analysis\n\nAs part of a larger programme of implementation research, we have previously developed an approach to evaluating and piloting essential interventions for NCDs in low-resource settings, and have adapted this methodology to suit the context of Tajikistan19.\n\nThe general approach to developing, piloting, and evaluating essential interventions for the management of hypertension and prevention of CVD in PHC in Tajikistan are outlined in the following seven steps. A more detailed methodological description is presented in the section that follows.\n\nStep One: Refresh clinical decision-making tools. Tajikistan national guidelines for PHC exist in a printed book format, and these books are available and used by family doctors. The national guidelines for hypertension (updated in 2018) endorse the use of WHO/ISH CVD risk charts, and some family doctors were previously trained on the use of WHO PEN. However, there is a lack of simple clinical decision support tools, such as laminated one-page algorithm cards for the treatment of hypertension. As such, the following tools will be adapted to be in-line with existing national guidelines, used for training, and disseminated to PHC workers in intervention clinics.\n\nHEARTS Hypertension Protocol – Diuretic as First-line Treatment\n\nHEARTS WHO PEN Diabetes Protocol\n\nWHO PEN Protocol 1 – adapted to integrate with HEARTS Hypertension Protocol\n\nWHO PEN Protocol 2 – adapted to integrated with HEARTS Hypertension Protocol\n\nBody-Mass Index Calculation Table\n\nWHO CVD Risk Charts – WHO/ISH no cholesterol charts\n\nIntegrated Management of Hypertension and Diabetes Algorithm (HEARTS Risk Module)\n\nAUDIT and Fagerstrom Clinical Decision Support Tools\n\nWhile the content of these tools is consistent with national guidelines, a ministerial order may be required to inform clinicians in the intervention clinics that it is appropriate to follow these updated versions of the already approved national guidelines.\n\nStep Two: Update training package for primary health care workers. There is an existing two-day training developed for the implementation of WHO PEN in Tajikistan. This training package will be reviewed for overall content, pre- and post-training evaluation of participant knowledge, adult learning methods, practical exercises, consistency with refreshed clinical decision support tools and protocols, and assessment of practical clinical skills.\n\nFurthermore, based on the previous experience of implementing WHO PEN in Tajikistan, the following topics have come to light as important to emphasize in the updated training package:\n\nOrganization of care\n\nTask sharing between doctors and nurses\n\nMeaningful clinical record keeping\n\nApproach to managing a patient without laboratory cholesterol tests\n\nAppropriate and timely follow-up of patients\n\nOptimization of nurse home-visit tasks to include greater focus on underserved populations at risk for CVD (e.g. older men)\n\nMeasuring inequalities in care and outcomes by gender\n\nRaising population awareness of cardiovascular diseases\n\nStep Three: Collection of baseline data. Baseline data will be collected from all clinics in the included district before implementation (i.e. training) begins. Data for quantitative indicators will be extracted by randomly sampling individual paper-based patient records from all PHC units using a standardized data collection instrument. This will be done by a small team of specially-trained national experts. After analysis, each health clinic will receive a summary report of their facility’s performance with respect to key project indicators. This report will not be powered for statistical significance, but will be used as a quality improvement exercise to show how routine clinical data can be used to improve clinical practice, for example by setting improvement targets.\n\nStep Four: Training staff in intervention clinics. All doctors and nurses from the PHC centres in the intervention arm will be invited to be trained together by a national team of experts in groups of approximately 30. It is estimated that up to 100 health workers will be trained in total. Given the human resources, the trainers will be themselves a group of family doctors and nurses, rather than narrow medical specialists or academics, so as to maintain direct relevancy to practical aspects of PHC services.\n\nThe training will be two-days; therefore it is estimated that two trainings can be conducted in one week (Monday/Tuesday and Thursday/Friday). Thus, 60 participants can be trained per week; two weeks will be required to train all doctors and nurses in the pilot district. A ministerial order will be required to relieve the health staff from their duties to attend the training.\n\nStep Five: Implementation of protocols and implementation coaching. Trained participants from intervention clinics will then be free to implement the clinical protocols and change their clinical practice, without incentives, for 12 months. During this time, a team of national and district experts will be created to offer support (distance and on-the-job) to the PHC centres. These teams will visit each clinic once every three months (for a total of four visits) and use a set of standardized instruments to review performance and provide constructive and timely feedback to the health facility manager and individual clinicians. The tools used during the support visits and implementation coaching will include:\n\nHEARTS Treatment Supervision Form – with minor adaptations to local context\n\nHEARTS Patient Interview Report Card – with minor adaptations to local context\n\nHEARTS Summary of Supervision Visits – with minor adaptations to local context\n\nStructured Interview and Debrief Conversation Template – to be developed\n\nIn addition to clinic visits every three months, representatives from each clinic will be brought together after 6 months of implementation for a workshop. This will be an opportunity for the project management team to provide support based on the needs identified in clinic visits, to encourage target setting, quality improvement and peer support, as well as gather more information about implementation barriers.\n\nStep Six: Collection of follow-up data. After 12 months, using the same method and data collection instruments used to collect baseline quantitative data (Step Three), data will again be extracted from randomly selected individual paper-based patient records from both intervention and control health centres. As was done at baseline, each health clinic will receive a summary report of their facility’s performance with respect to key project indicators comparing baseline to follow-up.\n\nOne-on-one semi structured interviews will be conducted with doctors, nurses, and health facility managers in the intervention clinics at 12 months follow-up, and these qualitative data will be analysed thematically for explanatory themes.\n\nStep Seven: Evaluation of results and sharing experience. The findings of the quantitative and qualitative analyses will be integrated in a final report and shared with key stakeholders, including health staff from the participating primary health care centres. The results will also be shared at a national conference and in an open-access peer reviewed journal, in order to inform the future development of primary health care capacity of Tajikistan and other similar settings.\n\nA pragmatic, sequential mixed methods explanatory design, composed of quantitative and qualitative strands will be used (Figure 1). The quantitative strand will be weighted more than the qualitative strand. A mixed methods design was chosen because it allows for the use of qualitative data to enlighten and explain the quantitative findings, but due to limited capacity to conduct qualitative research in Tajikistan, the qualitative strand will be weighted less than the quantitative strand.\n\nDue to resource constraints, and several ongoing PHC reform projects, a single district was nominated by the Ministry of Health and chosen for implementation (rather than multiple districts). All PHC in the district that meet inclusion criteria will be included; half will be randomly assigned to the intervention arm and half to the control arm. Baseline data will be collected from all clinics in the intervention arm and the control arm using the same standardized data extraction form.\n\nThe national working group initially nominated four districts for consideration which had participated in other projects: while three of these are subordinate districts (“Rayons”) of Dushanbe, the 4th (Vakhsh) is nearby to Dushanbe in the Oblast of Khatlon (Figure 2).\n\nIn order to provide more intensive support and ensure the quality of data collection, it was decided to focus on only one district. All PHC centres and the associated family doctors and nurses will be invited to take part in the training and pilot implementation. Shahrinav was chosen after considering its size, exposure to previous interventions, proximity to Dushanbe, year-round road access, and by recommendation from the Ministry of Health.\n\nWithin Shahrinav district, all PHCs were compared against the following exclusion criteria: (1) City Health Centre designation; (2) presence of narrow specialists (i.e. any physicians with specialization other than family medicine) within the clinic; (3) clinics with a family doctor to patient ratio of 1:4000 or more; or (4) clinics where information on the number of family doctors and/or family doctor to patient ratio is not available. If one or more of these four exclusion criteria were met, the clinic will not be included in the intervention or control arm.\n\nClinics allocated to the control arm will not receive the intervention; clinicians will proceed with usual care. The performance of individual health facilities will be compared from baseline to follow-up. To compare change in performance between intervention and control arms, the data from all intervention clinics will be aggregated together and compared to aggregated data of all control clinics.\n\nGiven the emphasis of hypertension management, including total CVD risk approach, indicators were developed to be calculated from a sample of hypertensive patients. These include indicators along the hypertension care pathway, including detection of hypertension, CVD risk assessment, risk factor modifying prescriptions, and BP control. Table 1 shows the primary and secondary indicators, the question the indicator seeks to answer, and the respective numerator and denominator definitions which will be used in the calculations.\n\nQuantitative data collection tool. We developed a standardized data collection tool to collect the anonymized data required for calculation of each indicator (Table 2). This tool will also be digitized such that it can be used on a computer (offline) or smartphone (online). We estimate it take 15 minutes per unique health record in order to extract data. Since there is a lack of electricity and mobile phone reception in many of the clinics, it is likely that data will have to be extracted first on paper and then later input to a database using an online data form. Data extractors will be blinded to the allocation of health care facilities to intervention or control for baseline data collection, but blinding is not possible for follow-up data collection because after the intervention period, allocation will be apparent when data collectors visit clinics and look at patient records.\n\nMethod of randomly sampling patient records. Each family doctor owns and maintains a single register (list of patients) for hypertensive patients (henceforth “hypertension register”). The hypertension register will be used to randomly select patient records. A computerized random number generator will be used to randomly select patients from the register. The selected patient chart will then be found and checked to see if it meets two inclusion criteria: (1) the patient must have visited the clinic at least once in the previous 12 months; and (2) the patient must have been 18 years or older 12 months prior to the date of data selection. If both criteria are met, the record will be used for data extraction. If the record does not meet inclusion criteria, it will be returned and a new record will be randomly selected. This process will be repeated until the sample size for each family doctor has been met (Figure 3).\n\nSample size. A total of 400 patient records will be sampled at baseline and follow-up from each of the intervention and control arms (n=800). The sample is based on a power calculation for the primary indicator (proportion of hypertensive patients whose blood pressure is controlled), estimated based on a WHO report that the control level in low and middle income countries is generally about 20–25%13. A 10% point difference between the intervention and control arms could be detected having 310 to 350 observations in each group using 0.05 type I error rate and 0.2 type II error rate. Sampling will be stratified by gender and selected in a 1:1 ratio of men to women (200 men and 200 women per arm). In the event that there are too few patients of a given gender, patients from the opposite gender will be substituted.\n\nData analysis. The change in indicators from baseline to follow-up will be calculated. Subgroup analysis by age, gender, and other demographic features may be done as deemed appropriate.\n\nQualitative data collection.\n\nSemi-structured interviews A purposive maximum variation sample of doctors and nurses from the intervention clinics will be invited to take part in semi-structured interviews at the end of the 12-month implementation phase. All participants will be required to provide written informed consent. Using an interview guide that will be adapted from previously published work21,22, interviews will be conducted one-on-one by members of the research team, audio recorded, and transcribed verbatim. Transcripts will be analysed thematically using framework thematic analysis, by two or more members of the research team, with oversight from an experienced qualitative researcher23.\n\nPatient and public involvement Neither patients nor the public were involved in the methodological design.\n\n\nStrengths and limitations of this study\n\nWe report methodology to adapt, pilot, and evaluate the WHO HEARTS technical package in a low-resource setting in a low-income country that can be used as an example for other jurisdictions with a high burden of cardiovascular diseases\n\nOur methods focus on evaluating the effect of the WHO HEARTS interventions in a real-world health system and therefore are expected to yield practical information to inform the national scale up of essential interventions for non-communicable diseases\n\nThe sample size is based on one primary indicator and it is therefore possible that some of the secondary indicators will be statistically underpowered\n\nDue to the practical and resource constraints, data collectors could not be blinded to follow-up data collection, and this could potentially bias the effect estimate\n\nAlthough an important stakeholder in implementation, patient perspectives were not included in the evaluation design\n\n\nCurrent study status\n\nData collection has been completed and analysis is underway.\n\n\nEthics and dissemination\n\nThis project has been reviewed and granted ethical approval from the Republican Clinical Center for Family Medicine under the Ministry of Health and Social Protection of the Republic Tajikistan. As per the ethical approval review and the nature of this project, consent was not obtained from individual patients for their anonymized routine clinical data to be used.\n\nFindings will be shared at the participant level (intervention clinics), national level through a national conference of key stakeholders, and internationally through publication in an open-access peer review journal. At the national level, the results will be used to assess the appropriateness and feasibility of national scale-up and mainstreaming into clinical practice. The raw data used in this project will not be made publicly available.\n\n\nData availability\n\nNo data are associated with this article.", "appendix": "References\n\nBenziger CP, Roth GA, Moran AE: The Global Burden of Disease Study and the Preventable Burden of NCD. Glob Heart. 2016; 11(4): 393–7. PubMed Abstract \| Publisher Full Text\n\nWorld Health Organization (WHO): World Health Statistics. 2018. Reference Source\n\nWorld Health Organization (WHO): The World Health Report 2008 - Primary Health Care (Now More than Ever). Geneva; 2008. Reference Source\n\nWorld Health Organization: The World Health Report - Health Systems Financing: the path to universal coverage. Geneva; 2010. Reference Source\n\nWorld Health Organization (WHO): Package of Essential Noncommunicable (PEN) disease interventions for primary health care in low-resource settings. 2013. Reference Source\n\nUnited Nations: Political Declaration of the High-Level Meeting of the General Assembly on the Prevention and Control of Noncommunicable Diseaes. Resoluation A/RES/66/2. New York; 2011. Reference Source\n\nWorld Health Organization (WHO): Global action plan for the prevention and control of NCDs 2013-2020. 2013. Reference Source\n\nWHO Regional Office for Europe: European Health Information Gateway. Copenhagen; 2018. Reference Source\n\nWorld Health Organization: Global Health Observatory - Total NCD Mortality by Country. Geneva; 2018. Reference Source\n\nWorld Health Organization: Risk of Premature Death from the Four Target NCDs. 2018. Reference Source\n\nWHO Regional Office for Europe: Understanding the health needs of Tajikistan. 2018. Reference Source\n\nKhodjamurodov G, Sodiqova D, Akkazieva B, et al.: Tajikistan: Health System Review. Health Syst Transit. 2016; 18(1): 1–114. PubMed Abstract\n\nMinistry of Health and Social Protection of the Population of the Republic of Tajikistan: National Health Strategy of the Republic of Tajikistan 2010-2020. Dushanbe; 2010. Reference Source\n\nWHO Regional Office for Europe: Review of the National Programme on the Development of Family Medicine 2011-2015 in Tajikistan. Copenhagen; 2016. Reference Source\n\nWorld Health Organization: Tackling NCDs: “Best buys” and other recommended interventions for the prevention and control of noncommunicable diseases. 2017. Reference Source\n\nWorld Health Organization: WHO/ISH Cardiovascular Risk Prediction Charts. Geneva; 2007. Reference Source\n\nWorld Health Organization: Global HEARTS Initiative. Geneva; 2018. Reference Source\n\nWorld Health Organization: HEARTS Technical Package. Geneva; 2018. Reference Source\n\nCollins D, Ciobanu A, Laatikainen T, et al.: Protocol for the evaluation of a pilot implementation of essential interventions for the prevention of cardiovascular diseases in primary healthcare in the Republic of Moldova. BMJ Open. 2019; 9(7): e025705. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJackson R, Ameratunga S, Broad J, et al.: The GATE frame: critical appraisal with pictures. ACP journal club. United States; 2006; 144(2): A8–11. Reference Source\n\nCollins DRJ, Jobanputra K, Frost T, et al.: Cardiovascular disease risk and prevention amongst Syrian refugees: mixed methods study of Médecins Sans Frontières programme in Jordan. Confl Health. 2017; 11: 14. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiew SM, Blacklock C, Hislop J, et al.: Cardiovascular risk scores: qualitative study of how primary care practitioners understand and use them. Br J Gen Pract. 2013; 63(611): e401–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGale NK, Heath G, Cameron E, et al.: Using the framework method for the analysis of qualitative data in multi-disciplinary health research. BMC Med Res Methodol. 2013; 13: 117. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "62985", "date": "15 May 2020", "name": "Anthony Etyang", "expertise": [ "Reviewer Expertise Epidemiology of hypertension and other non-communicable diseases in Africa" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nGeneral comments Dylan Collins and colleagues present a manuscript outlining the implementation and plans for evaluation of essential interventions for the management of hypertension and prevention of cardiovascular diseases in primary health care in Tajikistan. This is an important and timely piece of work and the authors are to be commended for a very well written manuscript. The comments that I give below are very minor issues that would in my opinion improve the manuscript. I also raise some points for the authors to consider about the actual project recognizing that it might be too late for them to implement them.\n\nSpecific Minor comments\nTitle of the manuscript- shouldn’t implementation precede evaluation? Evaluation is the seventh step in the processes outlined.\n\nAbstract- write out WHO PEN and HEARTS in full at first instance. 2nd last word of abstract-should read ‘reviewed’.\n\nDescription of health system in Tajikistan having a Soviet legacy in both the abstract and main text-: This appears inappropriate and unnecessary to me especially when juxtaposed against sentences describing need for out of pocket payments which point to a capitalist type of health system, similar for example to that of the USA. Cuba is a communist state and has a well functioning health system (please see additional comment on this below).\n\nP3, paragraph 4: Is the difference in the prevalence of obesity between men (15.4%) and women (11.9%) statistically significant? Please state if it is. I also find it unusual that the prevalence is higher in men than in women which is the opposite of the situation in most other countries. Are you able to offer any explanations for this?\n\nPlease explain what is meant by Soviet legacy in the health system. What specific aspects of the health system can be attributed to the fact that the country was part of the Soviet Union? As mentioned earlier, my feeling is that this reference to the past should be done away with unless a very clear and specific link can be shown.\n\nIt would be useful to provide background information on the numbers of health care workers in the country as well as Shahrinav district and the doctor/nurse to patient ratios.\n\nOverview of process and design section (p4). Step 1: Refresh clinical decision making tools- apart from having laminated one-page algorithm cards for clinicians to use in the treatment of hypertension, has consideration been made to convert these into smart phone applications? Is this feasible in Tajikistan?\n\nPlease include more detail on how the random allocation of clinics to control and intervention arms was/will be conducted. Were representatives of all the clinics present to observe the randomization taking place?\n\nTable 1: Is it possible to include magnitude of change in BP in mmHg as a secondary outcome? Proportion of controlled hypertensives is a rather crude measure.\n\nData analysis- This section needs to be expanded from the current two sentences as it will be critical to understanding the effect of the project. Is there a statistical analysis plan in place or is it being developed? For the primary outcome of proportion of hypertensives controlled, what test will be used and what confounders will be included in the model? Rather than comparing the change in indicators from baseline to follow up as the authors write, I would expect that the primary comparison would be proportion of controlled hypertension in intervention group versus proportion of controlled hypertension in control group at follow up. This is because the allocation to control versus intervention group is a randomized process and I would expect the proportions of controlled hypertensives in both trial arms to be essentially the same at baseline. It is important to have these statistical issues well thought out before data collection is completed.\n\nDissemination of findings: The authors should in the interests of open science consider having the data associated with the project made available to other researchers to reanalyze and build on their work. This can be through managed access or fully open processes.\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Not applicable", "responses": [] }, { "id": "66459", "date": "08 Sep 2020", "name": "Dagfinn Aune", "expertise": [ "Reviewer Expertise Epidemiology" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a study protocol for evaluation and pilot implementation of interventions for the management of hypertension and prevention of cardiovascular diseases in primary health care in Tajikistan. A pragmatic, sequential mixed methods explanatory design, composed of quantitative and qualitative strands will be used. A single geographic district was chosen for implementation and all primary care health care centers that meet inclusion criteria will be included. Half will be randomized to the intervention group and the other half to the control group. The overall process is organized in seven steps: refresh clinical decision-making tools – WHO PEN and HEARTS, update training package for health care workers, collection of baseline data, training staff in intervention clinics, implementation of protocols and implementation coaching, collection of follow-up data after 12 months, and evaluation of results and sharing experience.\n\nIt would have been interesting to know a bit more about the data that will be collected. Will you collect clinical measurements (weight, height, waist measures, blood pressure) and take blood samples (cholesterol, glucose etc.)? Will you collect and data on dietary intake, smoking, physical activity and other lifestyle factors? Will there be any repeated measurements during the study follow-up?\n\nWhat exactly will the intervention entail?\n\nI see now some of the information is in Table 2. You might want to have more detailed data on smoking – including never, former and current smoking, cigarettes per day and duration, time since quitting in former smokers.\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Partly\n\nAre the datasets clearly presented in a useable and accessible format? Not applicable", "responses": [] } ]	1	https://f1000research.com/articles/8-1639
https://f1000research.com/articles/8-1638/v1	13 Sep 19	{ "type": "Case Report", "title": "Case Report: Exploratory treatment with multiple intravenous infusion of the autologous adipose tissue-derived mesenchymal stem cells for the treatment of Diamond-Blackfan anemia patient", "authors": [ "Jeong Chan Ra", "Jong Im Park", "Pil Soon Yang", "Jong Im Park", "Pil Soon Yang" ], "abstract": "Diamond-Blackfan anemia (DBA) is a rare congenital erythropoietic disorder characterized by erythroblastopenia. Conventional treatments of DBA are the administrations of corticosteroids and blood transfusions for mitigation of anemia, and bone marrow transplantation. However, there are hurdles to overcome for long-term use and the conventional treatment. Mesenchymal stem cells (MSCs) have been noted as a novel alternative cell therapy in various diseases, and adipose tissue-derived MSCs (AdMSCs) are known for their versatile efficacies and feasibility. Here, we report the potential efficacies and the safety of intravenous administration of the autologous AdMSC in a patient with DBA for the first time. The isolation and characterization of autologous AdMSCs from a girl aged 11 years, 10 months with DBA were carried out due to the mutation of ribosomal protein s24 (RPS24). AdMSCs, diluted to 1 x 108 cells in 100 ml of saline, were infused intravenously for 1 hour. Intravenous administration of AdMSCs was carried out 5 times in 2-week intervals, and the patient was checked using various assessments (vital signs, physical examination, laboratory tests, adverse events, etc) at every visit. After 3, 6 and 9 months from the first administration of AdMSCs, red blood cell (RBC) count, hemoglobin value, and hematocrit were assessed for the efficacy. There were no side effects or adverse events observed during the treatment. Although showing subnormal values, the RBC number, hemoglobin level, and hematocrit were improved 9 months after the systemic administration of AdMSCs from baseline; the RBC count (x106 /μl), hemoglobin level (g/dl) and hematocrit level (%) were increased from 1.58 to 2.38, 5.6 to 8.3, and 16.9 to 26.1, respectively. The present case reported the first AdMSC administration for DBA patient and indicates it is possible that the intravenous administration of autologous AdMSC can be a safe alternative for DBA treatment.", "keywords": [ "Diamond-Blackfan anemia", "adipose tissue-derived mesenchymal stem cell", "intravenous injection" ], "content": "Introduction\n\nDiamond-Blackfan anemia (DBA) is a rare congenital bone marrow disorder, which characterized by erythroblastopenia1. The symptoms of DBA are generally presented in infancy, with broad congenital abnormality including growth retardation, defect of heart and urogenital system, malformation of hands, and cleft lip/palate2. It has been known that a mutation in the genes coding ribosomal proteins (RP) is responsible for DBA2. The large proportion of DBA patients have mutations detected in RP genes of the small (RPS7, RPS10, RPS15A, RPS17, RPS19, RPS24, RPS26, RPS27, RPS28, RPS29) or the large (RPL5, RPL11, RPL15, RPL18, RPL26, RPL27, RPL31, RPL35, RPL35A) ribosomal subunit3. In rare cases of DBA, two pathogenic X-linked mutations (GATA1, TSR2) were reported recently2. These ribosomal abnormalities in DBA are known to lead to the impairment of erythropoiesis4.\n\nThe treatment of DBA has conventionally focused on the mitigation of anemia resulting from impaired erythropoiesis. The administration of corticosteroids and blood transfusions have been commonly adopted for DBA patients2. However, steroids have various adverse effects after long-term use and the life-long transfusions for the patients who do not respond to steroids can also lead to problems, such as iron accumulation3. Allogenic hematopoietic stem cell (HSC) transplantation is considered a more fundamental approach for the steroid-refractory patients and to avoid iron accumulation from long-term transfusion5. Currently, the bone marrow from HLA-matching donors or cord blood transplantation have been carried out, and the positive outcomes have been reported6. However, allogenic HSC transplantation has considerable difficulties, such as the recruiting of HLA-matching donors, the in vitro propagation of HSCs, and the various complications from immunosuppression therapy, graft versus host disease (GVHD) etc.3.\n\nMesenchymal stem cells (MSCs) have been noted for a novel alternative in cell therapy in various diseases over past decades7,8. MSCs have been known for the effective tissue repair, angiogenesis, and the stimulation/mobilization of endogenous stem cells9–11. Adipose tissue-derived MSCs (AdMSCs) have shown several advantages over MSCs from other origins in safety and effectiveness, including a less invasive sampling procedure, abundant cell populations from tissue harvested and high capacity of proliferation12,13. Here we assessed the potential efficacies and the safety of intravenous administration of autologous AdMSCs in patients with DBA for the first time.\n\n\nCase presentation\n\nA girl aged 11 years and 10 months was admitted to our hospital (Bethesda Hospital, Yangsan, Korea). She presented with severe anemia, left blepharoptosis, and polydactylism in the right thumb at birth, and was diagnosed with pure red cell aplasia and congenital dyserythropoietic anemia in infancy. After seven transfusions in the first year of life, prednisolone or methylprednisolone were orally administrated at 1 mg/kg/day for 4 years from the age of 2.5 years. The patient was finally diagnosed with DBA due to a mutation in RPS24 at the age of 8 years by genetic analysis. The immediate family history was not identified. At that time of the visiting of the hospital for the treatment of AdMSC, the patient had already undergone four surgeries on her left blepharoptosis and had mild ptosis of the right eye; polydactylism in right hands was corrected after the surgery. The administration of autologous AdMSCs for DBA was approved by the Korean Ministry of Food and Drug Safety with Investigational New Drug Application for Emergency Use (Approval No. 20180101335). Written informed consent was acquired from the patient and patient’s parents before the initiation of treatment. This study was conducted in compliance with the Helsinki declaration and approved by Institutional Review Board of Bethesda Hospital (Approval No. 2018-FEB01).\n\nThe isolation and characterization of the autologous AdMSCs were performed using the previously established culture protocol14 under good manufacturing practice (GMP) conditions in the Stem Cell Research Institute of R Bio (Seoul, Republic of Korea). Briefly, the abdominal subcutaneous adipose tissue was obtained through liposuction, and digested with collagenase I (Gibco/Life Technologies, Grand Island, NY, USA). After centrifugation, the pellet was resuspended in DMEM (Invitrogen, Carlsbad, CA, USA)-based media containing 0.2 mM ascorbic acid and 10% fetal bovine serum (FBS; JR Scientific, Woodland, CA, USA). The cell fraction was cultured overnight at 37°C/5% CO2, and the cell adhesion was checked after 24 h. The cells were maintained for 4 to 5 days until confluent (passage 0). When the cells reached 90% confluency, they were subculture expanded in keratinocyte SFM-based media (Invitrogen, USA) containing 0.2 mM ascorbic acid, 0.09 mM calcium, 5 ng/ml rEGF, and 5% FBS until passage 3. Before transporting the cells for administration, aliquots of the AdMSCs are tested for cell viability, fungal, bacterial, endotoxin, and mycoplasma contamination and immunophenotype for MSCs. Cell viability evaluated by trypan blue exclusion was >91%, and no evidence of bacterial, fungal and mycoplasmal contamination was observed. The AdMSCs showed a homogenous population of cells with high positive marker expression levels of CD73 and CD90 at a high level of >85% and >99%, respectively. Negative markers of CD31, CD34, and CD45 were expressed at a very low level of <0.6%. AdMSCs finally adjusted to 1 × 108 cells in 100 ml of saline were injected intravenously for 1 hour. The intravenous administrations of AdMSC were carried out five times in 2-week intervals (June 21–Aug. 14, 2018), and the patient was instructed to visit the hospital to check the various assessments for safety (vital signs, physical examination, laboratory tests, adverse events, and serious adverse events).\n\nThe treatment schedule is showed in Figure 1. After 3 and 6 months from the first administration of AdMSCs, red blood cell (RBC) count, the hemoglobin value, hematocrit, reticulocyte percentage were assessed for the efficacy (Table 1).\n\nIV, intravenous; F/U, Follow-up; M, months; W, weeks.\n\n* The 2017 Korean National Growth Charts for children and adolescents\n\nThere were no side effects or adverse events observed during the administration of the autologous AdMSC. Most criteria of the assessment for the vital signs, physical examination, and laboratory tests showed the values within normal or clinically meaningless subnormal range. There was also no concomitant medication during the treatment of AdMSC. Although showing subnormal values, the number of RBC, hemoglobin level, and hematocrit level were improved after the systemic administration of AdMSC. The patient showed steady improvement in anemia after the treatment of AdMSCs; the RBC count was 2.38 ×106/μl hemoglobin level was 8.3 g/dl and hematocrit level was 26.1% after 9 months from the first administration of AdMSC (Table 1).\n\n\nDiscussion\n\nDBA is a congenital hematopoietic disorder caused by the mutation in genes of the ribosomal protein1. The mutation of RPS19 is founded in the approximate 25% of DBA cases, and RPS24 in 2% of cases1,4,15. The defects of RPS19, PRS24, and other genes involved in ribosomal biogenesis result in the impairment of the cell cycle and the protein synthesis rather than the differentiation process of the erythropoiesis, and finally give rise to the depletion of the erythrogenic progenitors16,17. Recently, gene therapy has been recommended as an emerging treatment, which could remedy the shortcomings of conventional treatment and be the fundamental cure3. However, the collection of genetically reprogrammed cells (induced pluripotent stem cells, HSCs, adult fibroblast, etc.) shows very low efficiency, and the risk of mutation limits the collecting of the clinical data and the application of treatment of humans3,4. AdMSCs, with its feasibility of the application, has already been identified as having potential for the treatment of various diseases, from chronic degenerative conditions to congenital defects or orphan diseases14,18. Autologous MSCs from patients with genetically defective disease could alleviate the patients’ own condition. The previous studies demonstrated that the autologous MSC from the patients with sickle cell anemia19 and autosomal dominant polycystic kidney disease20 were functionally compatible, and could be effective in the patients’ own status. The mechanisms of these efficacies of MSC are known to largely depend on the paracrine activity, which secretes the extracellular vesicles (exosomes) containing various cytokines, miRNAs, growth factors, and endocrine factors involving the bone regeneration, angiogenesis, immunomodulation, cellular proliferation, differentiation, recruiting of the endogenous stem cells21,22. Although the parameters were still in the subnormal range in the present case, the patient with DBA, who presented with severe anemia, showed a stable improvement from the anemic state after the intravenous administration of AdMSCs, without any other medications. The erythroid cells from patients with DBA have been reported to show defective expansion and proliferative properties23. However, the DBA erythroid colony formation could be enhanced under the presence of the exogenous stem cell factor (SCF, a KIT ligand) in vitro23–25. SCF is a major hematopoietic factor; the stromal cells from patients with DBA normally express SCF mRNA transcripts26. SCF secreted from the arterial endothelial cell in the bone marrow is essential for the intrinsic HSC27. AdMSCs have superb angiogenetic properties, which are considered to secrete the exosomes containing SCF, colony stimulating factor, interleukins and the other hematopoietic factors18,27,28. Further, AdMSCs stimulate the regeneration of endothelial cells through the paracrine angiogenetic factors, such as vascular endothelial growth factor, basic fibroblast growth factor, etc.22. In this regard, the encouraging outcomes in the present case can be assumed that there would be the direct hematopoietic supports from the patient’s own AdMSCs, and/or indirect hematopoietic help from the endothelial cells induced by the patient’ own AdMSCs. However, due to the restrictive case number, the lack of assessment of the extent of angiogenesis and the hematopoietic analysis after treatment with AdMSCs, there is a limitation to evaluate the potential of AdMSCs for erythropoiesis. Nonetheless, the present case report dealt on the first patient’ own AdMSCs administration for DBA treatment and gave the possibility that the intravenous administration of autologous AdMSCs can be a safe alternative for DBA patients.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\n\nConsent\n\nWritten informed consent for publication of the patients’ clinical details was obtained from the parents of the patient.", "appendix": "References\n\nSieff CA, Yang J, Merida-Long L, et al.: Pathogenesis of the erythroid failure in Diamond Blackfan anaemia. Br J Haematol. 2010; 148(4): 611–22. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDa Costa L, Narla A, Mohandas N: An update on the pathogenesis and diagnosis of Diamond-Blackfan anemia [version 1; peer review: 2 approved]. F1000Res. 2018; 7: pii: F1000 Faculty Rev-1350. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAspesi A, Borsotti C, Follenzi A: Emerging Therapeutic Approaches for Diamond Blackfan Anemia. Curr Gene Ther. 2018; 18(6): 327–335. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAspesi A, Betti M, Sculco M, et al.: A functional assay for the clinical annotation of genetic variants of uncertain significance in Diamond-Blackfan anemia. Hum Mutat. 2018; 39(8): 1102–1111. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRoggero S, Quarello P, Vinciguerra T, et al.: Severe iron overload in Blackfan-Diamond anemia: a case-control study. Am J Hematol. 2009; 84(11): 729–32. PubMed Abstract \| Publisher Full Text\n\nLi H, Lodish HF, Sieff CA: Critical Issues in Diamond-Blackfan Anemia and Prospects for Novel Treatment. Hematol Oncol Clin North Am. 2018; 32(4): 701–12. PubMed Abstract \| Publisher Full Text\n\nBrown C, McKee C, Bakshi S, et al.: Mesenchymal stem cells: Cell therapy and regeneration potential. J Tissue Eng Regen Med. 2019. PubMed Abstract \| Publisher Full Text\n\nTorres-Torrillas M, Rubio M, Damia E, et al.: Adipose-Derived Mesenchymal Stem Cells: A Promising Tool in the Treatment of Musculoskeletal Diseases. Int J Mol Sci. 2019; 20(12): pii: E3105. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShin L, Peterson DA: Human mesenchymal stem cell grafts enhance normal and impaired wound healing by recruiting existing endogenous tissue stem/progenitor cells. Stem Cells Transl Med. 2013; 2(1): 33–42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang M, Yuan Q, Xie L: Mesenchymal Stem Cell-Based Immunomodulation: Properties and Clinical Application. Stem Cells Int. 2018; 3057624. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAyala-Cuellar AP, Kang JH, Jeung EB, et al.: Roles of Mesenchymal Stem Cells in Tissue Regeneration and Immunomodulation. Biomol Ther (Seoul). 2019; 27(1): 25–33. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDufrane D: Impact of Age on Human Adipose Stem Cells for Bone Tissue Engineering. Cell Transplant. 2017; 26(9): 1496–1504. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMazini L, Rochette L, Amine M, et al.: Regenerative Capacity of Adipose Derived Stem Cells (ADSCs), Comparison with Mesenchymal Stem Cells (MSCs). Int J Mol Sci. 2019; 20(10): pii: E2523. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRa JC, Jeong EC, Kang SK, et al.: A Prospective, Nonrandomized, no Placebo-Controlled, Phase I/II Clinical Trial Assessing the Safety and Efficacy of Intramuscular Injection of Autologous Adipose Tissue-Derived Mesenchymal Stem Cells in Patients With Severe Buerger's Disease. Cell Med. 2016; 9(3): 87–102. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGazda HT, Grabowska A, Merida-Long LB, et al.: Ribosomal protein S24 gene is mutated in Diamond-Blackfan anemia. Am J Hum Genet. 2006; 79(6): 1110–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBadhai J, Fröjmark AS, J Davey E, et al.: Ribosomal protein S19 and S24 insufficiency cause distinct cell cycle defects in Diamond-Blackfan anemia. Biochim Biophys Acta. 2009; 1792(10): 1036–42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSong B, Zhang Q, Zhang Z, et al.: Systematic transcriptome analysis of the zebrafish model of diamond-blackfan anemia induced by RPS24 deficiency. BMC Genomics. 2014; 15(1): 759. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nŞovrea A, Boşca A, Constantin AM, et al.: State of the art in human adipose stem cells and their role in therapy. Rom J Morphol Embryol. 2019; 60(1): 7–31. PubMed Abstract\n\nStenger EO, Chinnadurai R, Yuan S, et al.: Bone Marrow-Derived Mesenchymal Stromal Cells from Patients with Sickle Cell Disease Display Intact Functionality. Biol Blood Marrow Transplant. 2017; 23(5): 736–45. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMakhlough A, Shekarchian S, Moghadasali R, et al.: Safety and tolerability of autologous bone marrow mesenchymal stromal cells in ADPKD patients. Stem Cell Res Ther. 2017; 8(1): 116. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBaraniak PR, McDevitt TC: Stem cell paracrine actions and tissue regeneration. Regen Med. 2010; 5(1): 121–43. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBaez-Jurado E, Hidalgo-Lanussa O, Barrera-Bailón B, et al.: Secretome of Mesenchymal Stem Cells and Its Potential Protective Effects on Brain Pathologies. Mol Neurobiol. 2019; 1–26. PubMed Abstract \| Publisher Full Text\n\nSieff C, Guinan E: In vitro enhancement of erythropoiesis by steel factor in Diamond-Blackfan anemia and treatment of other congenital cytopenias with recombinant interleukin 3/granulocyte-macrophage colony stimulating factor. Stem Cells. 1993; 11 Suppl 2: 113–22. PubMed Abstract \| Publisher Full Text\n\nAbkowitz JL, Sabo KM, Nakamoto B, et al.: Diamond-blackfan anemia: in vitro response of erythroid progenitors to the ligand for c-kit. Blood. 1991; 78(9): 2198–2202. PubMed Abstract\n\nOlivieri NE, Grunberger T, Ben-David Y, et al.: Diamond-Blackfan anemia: heterogenous response of hematopoietic progenitor cells in vitro to the protein product of the Steel locus. Blood. 1991; 78(9): 2211–5. PubMed Abstract\n\nSieff CA, Yokoyama CT, Zsebo KM, et al.: The production of steel factor mRNA in Diamond-Blackfan anaemia long-term cultures and interactions of steel factor with erythropoietin and interleukin-3. Br J Haematol. 1992; 82(4): 640–7. PubMed Abstract \| Publisher Full Text\n\nXu C, Gao X, Wei Q, et al.: Stem cell factor is selectively secreted by arterial endothelial cells in bone marrow. Nat Commun. 2018; 9(1): 2449. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRibeiro CA, Fraga JS, Grãos M, et al.: The secretome of stem cells isolated from the adipose tissue and Wharton jelly acts differently on central nervous system derived cell populations. Stem Cell Res Ther. 2012; 3(3): 18. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "98371", "date": "16 Nov 2021", "name": "Seth J Corey", "expertise": [ "Reviewer Expertise Molecular basis of inherited and congenital bone marrow failure syndromes" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this manuscript, the investigator report on the use of autologous mesenchymal stem cells for treatment of Diamond-Blackfan anemia. Following no preparative regimen and without any concomitant medication, 5 infusions of cultured adipocyte-derived MSC resulted in hemoglobin rise from 5.6 to 8.3 at 9 months. Retic from 1.3 to 1.9. No MCV was reported.\n\nThere are major concerns about this case report:\nNo detailing of MCV, white count, platelet count.\n\nNo detailing of lymphocyte subsets.\n\nNo rationale for use of autologous MSC. What is the hypothesized mechanism of action?\n\nDiamond-Blackfan anemia is a cell-autonomous disorder, without evidence of congenital hypoplastic anemia due to stromal disorder. Granted this individual had DBA secondary to RPS24 and had skeletal disorders — which might suggest some stroma/skeletal disorder. No discussion of bone marrow biopsies. Since this is a germline disorder, then even the stromal cells would have RPS24 mutation.\n\nBone marrow aspirates with culturing of mononuclear cells for BFU-E and CFU-E before and after MSC infusion are needed.\n\nHas this intervention (auto infusion of MSC) been done in other patients with Diamond-Blackfan anemia or other bone marrow failure syndromes.\n\nThis treatment was done August 2018. This manuscript was submitted 13 Sept 2019. This is Nov 2021. What has happened to the patient since then?\n\nIs the background of the case’s history and progression described in sufficient detail? No\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? No\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? No\n\nIs the case presented with sufficient detail to be useful for other practitioners? No", "responses": [] } ]	1	https://f1000research.com/articles/8-1638
https://f1000research.com/articles/8-616/v1	03 May 19	{ "type": "Research Article", "title": "Identification and phylogenetic analysis of oral Veillonella species isolated from the saliva of Japanese children", "authors": [ "Ariadna A. Djais", "Citra Fragrantia Theodorea", "Izumi Mashima", "Maiko Otomo", "Masato Saitoh", "Futoshi Nakazawa", "Citra Fragrantia Theodorea", "Izumi Mashima", "Maiko Otomo", "Masato Saitoh", "Futoshi Nakazawa" ], "abstract": "Background: As the most frequent infectious disease among children worldwide, dental caries have a strong relationship with oral hygiene status, specifically in the development of infection. Oral Veillonella species have a main role as early colonizers in the oral biofilm formation. Previously, oral Veillonella strains were detected at the species level in the saliva of Thai children with different oral hygiene statuses. Here, we studied the oral hygiene status by examining the composition and proportion of oral Veillonella species in saliva of Japanese children to compare with the previous results found in Thai children. Methods: Microbial samples collected from 15 Japanese children divided into three oral hygiene groups were cultured under anaerobic conditions after homogenization and dilution, and inoculated onto brain heart infusion and selective medium Veillonella agar. Genomic DNA was extracted from each isolate. Veillonella species were detected by one-step PCR using rpoB species-specific primers. To analyse the phylogenetic properties of the unknown Veillonella strains, PCR amplification and sequence analysis of rpoB were conducted for 10 representative strains. Results: Although V. rogosae was found as the predominant species among all groups, its prevalence was significantly lower in the children with poor oral hygiene than in those with good oral hygiene. V. parvula was the prevalent species in the poor oral hygiene group. Approximately 10% of the isolated Veillonella strains were not classified to any established species; the phylogenetic analysis showed that they were most closely related to V. infantium Conclusions: This study demonstrates that the composition and proportion of oral Veillonella species in the saliva of Japanese children is correlated with different oral hygiene status. Changes in detection ratios of V. parvula and V. rogosae can be useful indicators of oral hygiene status. Furthermore, new strains closely related to V. infantium were isolated from the saliva of Japanese children.", "keywords": [ "Oral Veillonella", "dental caries", "oral hygiene status", "indicator", "phylogenetic", "saliva", "children", "Japan" ], "content": "Introduction\n\nThe oral biofilm comprises multiple bacterial species and develops as a result of adhesion of pioneer bacterial species to adsorption of salivary proteins and glycoproteins on the enamel surface. These biofilms are not formed by random simultaneous colonization, but rather by selective, reproducible, and sequential colonization1,2. Oral biofilms are a source of bacteria that cause oral infections, for instance dental caries and periodontal disease, and they sometimes lead to or worsen systemic diseases3.\n\nSaliva is an acknowledged pool of biological markers that range from biochemical molecules changes such as DNA, RNA, and proteins, to those in microbiota structural composition4. Furthermore, saliva has an important role in oral biofilm development and maintenance. Recently, metagenomic analysis from saliva samples of Thai children demonstrated that Streptococcus and Veillonella were the predominant bacterial genera in the samples, and the proportion of Streptococcus decreased, while that of Veillonella increased in the children with poor oral hygiene status5.\n\nThe genus Veillonella consist of multiple gram-negative bacterial species, obligate anaerobic, non-motile, non-spore forming, small cocci belonging to the family Veillonellaceae6.\n\nNo Veillonella species ferment carbohydrates or amino acids, except for V. criceti, V. ratti, and V. seminalis. The metabolic end products of Veillonella species from trypticase-glucose-yeast extract are mainly acetic acid and propionic acid6. Veillonella species are present as commensal organisms in the oral cavity, intestinal tract and genitouritary and respiratory systems of humans and animals. Previous studies have reported that Veillonella species are rare causative organisms of meningitis, endocarditis, bacteraemia, discitis, vertebral osteomyelitis, and prosthetic joint infection7–9. Generally, Veillonella species are known to be resistant to tetracycline and sensitive to penicillin and ampicillin. However, some Veillonella strains resistant to both penicillin and ampicillin have recently emerged10.\n\nThere are 14 species reported to belong to genus Veillonella including V. infantium which was assign as a novel species in 201811. Of the 14 documented species, V. atypica, V. denticariosi, V. parvula, V. rogosae, V. dispar, V. infantium, and V. tobetsuensis have been found in human saliva or on tongue or dental biofilms12–17. Periasamy and Kolenbrander reported that oral Veillonella species are an early colonizer during the formation of oral biofilm, along with Streptococcus species, which were reported as initial colonizers in developing multispecies communities of oral biofilm18. Therefore, it is important to determine the role of oral Veillonella species in formation of oral biofilm to improve the prevention and treatment of oral infectious diseases.\n\nVeillonella strains are relatively easy to identify at the genus level, but remain difficult to idnetify at the species level, since there are no useful phenotypic or biochemical examinations to distinguish them19. To resolve this problem, Mashima et al. established a novel one-step PCR method with species-specific primer sets based on the variable region of the rpoB gene sequences of oral Veillonella species12. Additionally, 1,442 Veillonella strains isolated from the saliva of 107 Thai children were identified by this method as V. dispar, V. parvula, V. rogosae, V. atypica, V. denticariosi, and V. tobetsuensis in our previous study20. In that study, V. parvula was significantly more prevalent in the poor oral hygiene, and the detection rate of oral Veillonella species in the saliva was indicative of the oral hygiene status of Thai children20. Additionally, another study suggested that several novel Veillonella species may inhabit the human oral cavity21.\n\nTherefore, in this study, we examined composition and proportion of oral Veillonella species in saliva of Japanese children with different oral hygiene status. We assumed that the detection rate and distribution of oral Veillonella species in saliva detected in Japanese children were similar to those reported in Thai children.\n\nFurthermore, we determined the phylogenetic position of the unknown Veillonella strains evaluated by the genus-specific PCR primer set as members of the genus Veillonella with a phylogenetic tree.\n\n\nMethods\n\nThe 15 children selected to take part in the study were 6 boys and 9 girls, aged 4 to 14 years old. Participants were recruited in-person during appointments at the Dental Hospital, Health Sciences University of Hokkaido. The subjects who had a history of immunosuppression or systemic diseases (e.g. leukemia, hepatitis), or any conditions requiring antibiotic monitoring or treatment procedures (e.g. heart conditions, bone fractures), or those with mucosal lesions, previous chemotherapy, radiation therapy, or medications that can reduce the salivary flow, and those that underwent treatment with antimicrobials within the previous three months were excluded from this study.\n\nSubjects of this study were divided into three groups based on their evaluation by the Simplified Oral Hygiene Index (OHI-s) into good, moderate, and poor oral hygiene groups, according to the criteria of Greene and Vermillion22. Owing to the small number of children with poor hygiene (n=5), it was decided that 5 children would be chosen for each group. The good oral hygiene group (OHI-S score: 0–1.2) was composed of two males and three females. The moderate group (OHI-S score: 1.3–3.0) was composed of 3 males and 2 females. The poor group (OHI-S score: 3.1–6.0) was composed of 1 male and 4 females.\n\nThe saliva samples were collected at the Dental Hospital, Health Sciences University of Hokkaido, Japan, over a period between 2016 and 2017. Saliva was stimulated by paraffin chewing for 1 min and was then collected into sterile plastic tubes, and transferred to an anaerobic box (Hirasawa Works, Inc., Osaka, Japan) containing 10% H2, 85% N2, 5% CO2. These samples (1 ml each) were transferred to 1.5-ml Eppendorf tubes, then homogenized for 1 min with a BioMasher ®II (Nippi, Incoporated Protein Engineering Office, Tokyo, Japan).\n\nThese homogenized saliva samples were serially diluted by 10-fold with sterile phosphate buffer saline (PBS) from 10-3 to 10-7. Aliquots (100 µl) of each diluted sample were inoculated into BactoTM Brain Heart Infusion (BHI, Difco Laboratories, Detroit, MI, USA) supplemented with 5% (volume/volume) defibrinated sheep blood (BHI agar), hemin (10 μg/mL, Wako, Osaka, Japan), menadione (5 µg/ml, Wako), and the selective medium Veillonella agar23. After inoculation, all media were incubated under anaerobic conditions with 10% H2, 85% N2, and 5% CO2 at 37°C. Veillonella agar was incubated for 5 days and BHI agar was incubated for 7 days. The bacterial colonies grown on BHI and Veillonella agar were counted as the total number of bacteria and typical Veillonella colonies in the saliva sample, respectively. Bacterial cells of typical Veillonella colonies were confirmed as gram-negative cocci with light microscopy after gram staining. Standard strains consisted of V. atypica ATCC 17744T, V. denticariosi JCM 15641T, V. dispar ATCC 17748T, V. parvula ATCC 10790T, V. rogosa JCM 15642T, and V. tobetsuensis ATCCBAA-2400T.\n\nThe genomic DNA was extracted from the isolated bacterial cells by using Insta Gene Matrix Kit (Bio-Rad Laboratories, Hercules, CA, USA). The DNA concentration determination was based on fluorescence by using a Qubit 3.0 Fluorometer. (Invitrogen, Carlsbad, CA, USA), according to the manufacturer’s protocol. Additionally, genomic DNA extracted from the standard strains stated above was used as positive control for PCR.\n\nFor the identification of Veillonella species at the genus level, a genus-specific PCR primer pair, Veill-rpoBF and Veill-rpoBR, were used according to the protocols described by Arif et al. and Mashima et al.12,13,24. Strains confirmed by PCR as members of genus Veillonella were then subject to the one-step PCR method with the species-specific primers sets ATYR, DENR, DISR, PARR, ROGR, TOBR, and VF, performed according to the method reported by Mashima et al., for identification at species level12.\n\nThe PCR products were applied to a 2.0% agarose gel, and after electrophoresis, the gel was stained with SYBR® Safe DNA gel stain (Invitrogen).\n\nFor phylogenetic analysis of unknown strains, genomic DNA was also extracted from bacterial cells of unknown Veillonella strains showing positive PCR reaction with the genus-specific primer, but negative with the species-specific primer sets. In addition, PCR-based amplification and sequence analysis of rpoB were performed using the previously described primers for genus Veillonella rpoB-forward (5’-GTA ACA AAG GTG TCG TTT CTC G-3’) and rpoB-reverse (5’-GCA CCR TCA AAT ACA GGT GTA GC-3’)24.\n\nThe PCR product contained DNA fragments were purified by using QIAquick® Gel Extraction Kit (Qiagen, Hilden, NW, Germany), according to the manufacturer’s instructions. The DNA concentration after purification was determined based on fluorescence using a Qubit® 3.0 Fluorometer dsDNA HS Assay Kit (Invitrogen life Technologies, Carlsbad, CA, USA. The PCR reaction was performed with 15–20 ng/µl of DNA template for cycle sequence.\n\nPurified DNA from PCR was sequenced with an BigDye® Terminator v1.1 Cycle Sequencing kit (Thermo Fisher, Waltham, MA, USA), BigDye® Terminator 5X Sequencing Buffer (Thermo Fisher, Waltham, MA, USA), single primer 1 pM and PCR product in a final volume of 20 µl. Cycle sequencing of the purified DNA was as follows: preheating at 96°C for 1 minutes; followed by 25 cycles of denaturation at 96°C for 10 seconds and annealing with extension at 60°C for 4 minutes12. Furthermore, the sequencing of PCR products were purified by using Centri-Sep column (Princeton Separations, Adelphia, NJ, USA), according to the manufacture’s instruction and resolved for the sequencing analysis.\n\nDNA sequences were determined using an ABI PRISM 310 Genetic Analyzer (Applied Biosystem) and were aligned and connected using SEQMAN Pro from the LASERGENE program (DNASTAR). The programs MEGALIGN, which includes CLUSTALW and NJPlot were used to compare sequences and to reconstruct an evolutionary tree by the neighbour-joining method25. Confidence intervals were also assessed by CLUSTALW with bootstrap analysis. Furthermore, pairwise similarity values were determined with MEGALIGN in the LASERGENE program. The rpoB sequences of the unknown Veillonella strains were aligned against the sequence of the established Veillonella species retrieved from GenBank. Unipro UGENE could be use as free alternative for both sequencing and pairwise similarity values.\n\nAll subjects and their parents were made aware of the objectives and procedures of the study and parents of participants provided written informed consent. This study was conducted with the approval of The Ethics Committee of the Health Sciences University of Hokkaido, Japan, under process number of 2016-015\n\n\nResults\n\nThe average number of colony forming units (CFU)/ml of all bacteria on BHI agar increased with decreased oral hygiene: 1.38E+08, 2.2E+08 and 4.48E+09 in the good, moderate and poor groups, respectively. Raw CFU data are available on Figshare26.\n\nThe phenotypic characteristics of Veillonella colonies on the selective medium were 2–4 mm in diameter, and slightly domed in shape with an entire edge, opaque, and greyish white. They were composed of small, gram-negative coccal cells, mainly existing as single cells but with short chains visible. In the good oral hygiene group, a mean numberof 1.70E+06 CFU/ml (median, 1.20E+06 CFU/ml) were identified as the genus Veillonella, with 12.3% V. atypica, 19.3% V. dispar, 10.5% V. parvula, 49.1% V. rogosae, and 8.8% unknown species (Table 1). In the moderate group, 2.08E+07 CFU/ml with median 2.00E+06 were identified as the genus Veillonella, with 6.2% V. atypica, 29.6% V. dispar, 12.3% V. parvula, 44.4% V. rogosae, and 7.4% unknown species (Table 2). Meanwhile, in the poor oral hygiene group, 4.48E+09 CFU/ml with median 2.20E+06 were identified as the genus Veillonella, with 7.3% V. atypica, 12.2% V. dispar, 31.7% V. parvula, 34.1% V. rogosae, and 14.6% unknown species proportions (Table 3).\n\nThe colony-forming units (CFU) of all anaerobic bacteria on brain heart infusion agar and Veillonella strains on Veillonella agar (detection limit <0.1% of the total count). The total of Veillonella isolates identified by the Veillonella genus-specific PCR primer. Individual species as a percentage of the number of isolates identified by one-step PCR with the species-specific primer sets for each subject (n = 5) from saliva of the good oral hygiene group.\n\nThe colony-forming units (CFU) of all anaerobic bacteria on the brain heart infusion agar and Veillonella strains on Veillonella agar (detection limit <0.1% of the total count). The total of Veillonella isolates identified by the Veillonella genus-specific PCR primer. Individual species as a percentage of the number of isolates identified by one-step PCR with the species-specific primer sets for each subject (n = 5) from saliva of the moderate oral hygiene group.\n\nThe colony-forming units (CFU) of all anaerobic bacteria on the brain heart infusion agar and Veillonella strains on Veillonella agar (detection limit <0.1% of the total count). The total of Veillonella isolates identified by the Veillonella genus-specific PCR primer. Individual species as a percentage of the number of isolates identified by one-step PCR with the species-specific primer sets for each subject (n=5) from saliva of the poor oral hygiene group.\n\nAs shown in the results, V. rogosae was found as the predominant species in the saliva samples of all oral hygiene groups. However, V. denticariosi and V. tobetsuensis were not found in all oral hygiene groups (Table 1–Table 3). Figure 1 shows the per cent ratio of the total number of strains of each species to the total number of Veillonella isolates from saliva samples of the good, moderate, and poor oral hygiene groups.\n\nData expressed as percentage of total the total isolated number as Veillonella in samples from saliva in good, moderate and poor oral hygiene groups.\n\nIn this study, 179 strains were identified as Veillonella strains, and 162 strains were identified as V. atypica, V. dispar, V. parvula or V. rogosae. However, 17 (9.5%) of 179 strains failed to be classified as any of the known oral Veillonella species, thus, they were defined as unknown species. Of the 17 unknown Veillonella strains 10 (S3-1, S9-1, S10-1, S15-1, S17-1, S21-1, S25-2, S28-1, S29-1 and S30-1) were selected as representative strains from different hygiene groups for phylogenetic analysis. After determination of the rpoB sequences of these 10 strains, these sequences were aligned toward to the sequences of Veillonella type strains retrieved from GenBank. The evolutionary tree produced by analysing the type strains of the 14 Veillonella species and the 10 unknown strains is shown in the Figure 2. According to this data, the most closely related species was V. infantium, although the 10 unknown strains formed three clusters. The DNA sequence similarity of the 10 unknown Veillonella strains to V infantium JCM 31738T (LC191258) ranged from 97.1 to 99.7%.\n\nBootstrap values were indicated at corresponding nodes. The scale bar indicated the number of nucleotide substitutions per 100 residues.\n\n\nDiscussion\n\nIt was previously reported that a higher number of anaerobic bacteria was detected on BHI agar in saliva from Thai children with poor oral hygiene than those with good and moderate oral hygiene20. This prior study demonstrated that oral Veillonella isolates were detected at a twofold higher frequency in the saliva of Thai children with poor rather than good or moderate oral hygiene20. Here, it was demonstrated that the number of anaerobic bacteria on BHI agar and Veillonella species on the selective medium increased in saliva of Japanese children with worsening oral hygiene status. Therefore, the detection level of anaerobic bacterial strains and oral Veillonella strains in saliva from Japanese children with good, moderate and poor oral hygiene status was similar to that from Thai children.\n\nUsing the Illumina MiSeq platform, Mashima et al. demonstrated that Streptococcus and Veillonella species were the predominant bacterial species in the saliva microbiome of Thai children, but that the proportion of Streptococcus decreased while that of Veillonella increased with poor oral hygiene status5. They also found that Veillonella species were detected predominantly in the tongue microbiome of Thai children with poor oral hygiene status compared to those with good or moderate oral hygiene status5. Taken together with the results of the present study, it is possible that Veillonella species could be a biomarker of oral hygiene status for Thai and Japanese children.\n\nThis study demonstrated that V. rogosae was the predominant species detected in all groups of Japanese children (Figure 1). Beighton et al. reported V. rogosae as one of the predominant Veillonella species in tongue biofilms of healthy adults in the UK24. A previous study also showed that V. rogosae was the predominant Veillonella species in tongue biofilms of the children in Thailand12. Recently, Theodorea et al. isolated 1,609 Veillonella strains from saliva samples of Thai children divided into three groups: good, moderate and poor oral hygiene status20. Then, 1,442 of 1,609 strains were detected by the one-step PCR method with the species-specific primer sets for V. atypica, V. denticariosi, V. dispar, V. parvula, V. rogosae, or V. tobetsuensis. They reported that V. rogosae was the predominant species detected in all groups20. These results of the previous and present studies indicate that V. rogosae is the predominant oral Veillonella species in the human saliva and tongue biofilm.\n\nFurthermore, this study showed that the detection rate of V. rogosae decreased as oral hygiene quality decreased: 49.1%, 44.4%, and 34.1% in the good, moderate, and poor oral hygiene groups, respectively (Figure 1). Similar results were obtained from saliva samples of Thai children were also reported by Theodorea et al.20. Based on these results, it was demonstrated that the detection rate of V. rogosae decreased with aggravation of oral hygiene status in Japanese and Thai children. Additionally, Arif et al. detected V. rogosae only in carious-free lesions of dental plaques13. All these data suggest that a human oral cavity with good hygiene status is suitable habitat for V. rogosae.\n\nConversely, the detection rate of V. parvula in the poor (31.7%) oral hygiene was significantly higher than that in the good (10.5%) and moderate (12.3%) oral hygiene groups, in this study with Japanese children. This result is conforms with data from another study, in which V. parvula was most frequently detected in saliva of Thai children with poor oral hygiene status20. Previously, it was also reported that V. parvula was frequently detected in periodontal pockets27 and active carious-lesions28. These data suggest that oral cavities with poor hygiene status are suitable environments for V. parvula.\n\nIn this study, although 179 strains were isolated members of the genus Veillonella from saliva of 15 Japanese children, V. denticariosi and V. tobetsuensis were not found in any samples. In the case of saliva samples from Thai children, the detection rate of V. denticariosi (0.4%) and V. tobetsuensis (1.7%) were very low, although in this study 1,609 Veillonella strains were isolated from 107 Thai children20. Similarly, it was reported that V. denticariosi was not detected in any of the tongue biofilms of Thai children12, and V. denticariosi was detected in tongue biofilm of only one young Japanese adult17. Therefore, V. denticariosi may be the least prevalent oral Veillonella species in the saliva and tongue microbiome. On the other hand, V. tobetsuensis, was not detected in saliva from Thai children with good oral hygiene status. However, the detection rate of V. tobetsuensis was 14.3% and 17.8% in the saliva of Thai children with moderate and poor oral hygiene, respectively (20). Similarly, it was demonstrated that the prevalence of V. tobetsuensis ranged from 7.6% to 20.0% in tongue biofilm samples from Japanese adults16. Therefore, these data suggest that V. tobetsuensis may be potential to co-occur with other Veillonella species in saliva and tongue biofilms.\n\nIn the present study with saliva samples of Japanese children, 17 (9.5%) of 179 strains confirmed as member of genus Veillonella were not belong to any established Veillonella species as unknown species. Theodorea et al., also reported that 167 (10.4%) of 1,609 Veillonella isolates from saliva of Thai children could not be assigned to any species of the genus Veillonella20. Furthermore, it was reported that 43 (9.7%) of the 442 Veillonella isolates from periodontal pockets and gingival sulcus could not be identified as any of the known Veillonella species. These data may indicate that other novel Veillonella species inhabit human oral cavities, although only six species were detected as oral Veillonella species up to this point. Further phylogenetic studies are needed to evaluate the possibilities of novel Veillonella species.\n\nIn 2018, Mashima et al.11 proposed V. infantium as a novel species isolated from saliva of Thai children, representing a seventh oral Veillonella species. Therefore, for phylogenetic analysis of the unknown Veillonella strains isolated in this study, the rpoB sequences of type strains of the established Veillonella species, including V. infantium JCM 31738T, were examined. Consequently, although 10 unknown Veillonella strains analysed in this study formed three clusters distinct from V. dispar, the most closely related species was V. infantium. Further studies are required to assign these strains most accurately.\n\nIn conclusions, this is the first study to identify oral Veillonella at the species level in the saliva of Japanese children divided into three oral hygiene status groups: good, moderate and poor group. Although V. denticariosi and V. tobetsuensis were not found in any groups in this study because of small number of subjects, the distribution and frequency of V. atypica, V. dispar, V. parvula and V. rogosae, were mostly the same as those in the saliva from Thai children divided into the aforementioned oral hygiene status groups. Additionally, the results of this study demonstrate that changes in the ratio of some Veillonella species, such as an increase of V. parvula and decrease of V. rogosae in those with poor oral hygiene, can be a useful indicator of oral hygiene status, as with results obtained in the study of saliva taken from Thai children. The present study also showed that approximately 10% of the isolated Veillonella strains were not classified to any Veillonella species, and that they will be assigned to V. infantium or novel Veillonella species after further studies.\n\n\nData availability\n\n16S rRNA sequences of the 10 unknown Veillonella strains are available from GenBank, accession numbers: LC467206 (S9-1), LC467207 (S28-1), LC467208 (S25-1), LC467209 (S17-1), LC467210 (S15-1), LC467211 (S29-1), LC467212 (S10-1), LC467213 (S30-1), LC467214 (S21-1), LC467215 (S3-1).\n\nFigshare: Raw Data Saliva Japanese Children.xlsx. https://doi.org/10.6084/m9.figshare.7856657.v126.\n\nThis project contains the number of colony-forming units and total number of Veillonella strains isolated from each child.\n\nData on Figshare are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).", "appendix": "Grant information\n\nThis study was supported in part by grants-in aid from the Japan Society for the promotion of Science Fellows (15J30007), Scientific Research from KAKENHI (26462793), 2016-2017 Research Project of Institute of Personalized Health Sciences, Health Sciences University of Hokkaido, and 2016-2017 research grant from the Institute for Fermentation, Osaka, Japan.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nThe authors wish to thank all the children and their parents for their valuable participation in this study.\n\n\nReferences\n\nBjarnsholt T, Alhede M, Alhede M, et al.: The in vivo biofilm. Trends Microbiol. 2013; 21(9): 466–474. PubMed Abstract \| Publisher Full Text\n\nNyvad B, Kilian M: Microbiology of the early colonization of human enamel and root surfaces in vivo. Scand J Dent Res. 1987; 95(5): 369–380. PubMed Abstract \| Publisher Full Text\n\nMaddi A, Scannapieco FA: Oral biofilms, oral and periodontal infections, and systemic disease. Am J Dent. 2013; 26(5): 249–254. PubMed Abstract\n\nZhang CZ, Cheng XQ, Li JY, et al.: Saliva in the diagnosis of diseases. Int J Oral Sci. 2016; 8(3): 133–137. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMashima I, Theodorea CF, Thaweboon B, et al.: Exploring the salivary microbiome of children stratified by the oral hygiene index. PLoS One. 2017; 12(9): e0185274. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJean-Philippe C: Genus Veillonella Prèvot 1933, 118AL emend. Mays, Holdeman, Moore, Rogosa and Johnson 1982, 35. In Bergey’s Manual of Sysematic Bacteriology. Second Edition, Edited by Paul DV, George MG, Dorothy J, Noel RK, Wolfgang L, Fred AR, Karl-Heinz S & William BW. Springer: Dordrecht, Heidelberg, London & New York, 2009; 3: 1059–1065. Reference Source\n\nIsner-Horobeti ME, Lecocq J, Dupeyron A, et al.: Veillonella discitis. A case report. Joint Bone Spine. 2006; 73(1): 113–5. PubMed Abstract \| Publisher Full Text\n\nLiu JW, Wu JJ, Wang LR, et al.: Two fatal cases of Veillonella bacteremia. Eur J Clin Microbiol Infect Dis. 1998; 17(1): 62–4. PubMed Abstract \| Publisher Full Text\n\nMarriott D, Stark D, Harkness J: Veillonella parvula discitis and secondary bacteremia: a rare infection complicating endoscopy and colonoscopy? J Clin Microbiol. 2007; 45(2): 672–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nReady D, Bedi R, Mullany P, et al.: Penicillin and amoxicillin resistance in oral Veillonella spp. Int J Antimicrob Agents. 2012; 40(2): 188–9. PubMed Abstract \| Publisher Full Text\n\nMashima I, Liao YC, Miyakawa H, et al.: Veillonella infantium sp. nov., an anaerobic, Gram-stain-negative coccus isolated from tongue biofilm of a Thai child. Int J Syst Evol Microbiol. 2018; 68(4): 1101–1106. PubMed Abstract \| Publisher Full Text\n\nMashima I, Theodorea CF, Thaweboon B, et al.: Identification of Veillonella Species in the Tongue Biofilm by Using a Novel One-Step Polymerase Chain Reaction Method. PLoS One. 2016; 11(6): e0157516. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nArif N, Do T, Byun R, et al.: Veillonella rogosae sp. nov., an anaerobic, Gram-negative coccus isolated from dental plaque. Int J Syst Evol Microbiol. 2008; 58(Pt 3): 581–584. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nByun R, Carlier JP, Jacques NA, et al.: Veillonella denticariosi sp. nov., isolated from human carious dentine. Int J Syst Evol Microbiol. 2007; 57(Pt 12): 2844–2848. PubMed Abstract \| Publisher Full Text\n\nMashima I, Kamaguchi A, Miyakawa H, et al.: Veillonella tobetsuensis sp. nov., an anaerobic, gram-negative coccus isolated from human tongue biofilms. Int J Syst Evol Microbiol. 2013; 63(Pt 4): 1443–1449. PubMed Abstract \| Publisher Full Text\n\nMashima I, Nakazawa F: Identification of Veillonella tobetsuensis in tongue biofilm by using a species-specific primer pair. Anaerobe. 2013; 22: 77–81. PubMed Abstract \| Publisher Full Text\n\nMashima I, Kamaguchi A, Nakazawa F: The distribution and frequency of oral veillonella spp. in the tongue biofilm of healthy young adults. Curr Microbiol. 2011; 63(5): 403–407. PubMed Abstract \| Publisher Full Text\n\nPeriasamy S, Kolenbrander PE: Central role of the early colonizer Veillonella sp. in establishing multispecies biofilm communities with initial, middle, and late colonizers of enamel. J Bacteriol. 2010; 192(12): 2965–2972. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKolenbrander PE, Moore LVH: The genus Veillonella. In: The Prokaryotes. (2nd edn), Springer, New York, USA. 1992; 2034–2047. Publisher Full Text\n\nTheodorea CF, Mashima I, Thaweboon B, et al.: Molecular detection of oral Veillonella species in the saliva of children with different oral hygiene statuses. Int J of Curr Microbiol Appl Sci. 2017; 6(7): 449–461. Reference Source\n\nTheodorea CF, Mashima I, Nakazawa F: Prospects of novel species of oral Veillonella in human saliva. Adv Biotech Microbiol. 2017; 5(4): 1–3. Publisher Full Text\n\nGreene JC, Vermillion JR: The Simplified Oral Hygiene Index. J Am Dent Assoc. 1964; 68(1): 7–13. PubMed Abstract \| Publisher Full Text\n\nRogosa M, Fitzgerald RJ, Mackintosh ME, et al.: Improved medium for selective isolation of Veillonella. J Bacteriol. 1958; 76(4): 455–456. PubMed Abstract \| Free Full Text\n\nBeighton D, Clark D, Hanakuka B, et al.: The predominant cultivable Veillonella spp. of the tongue of healthy adults identified using rpoB sequencing. Oral Microbiol Immunol. 2008; 23(4): 344–347. PubMed Abstract \| Publisher Full Text\n\nSaitou N, Nei M: The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol. 1987; 4(4): 406–25. PubMed Abstract \| Publisher Full Text\n\nDjais AA: Raw Data Saliva Japanese Children.xlsx. figshare. Dataset. 2019. http://www.doi.org/10.6084/m9.figshare.7856657.v1\n\nMashima I, Fujita M, Nakatsuka Y, et al.: The Distribution and Frequency of Oral Veillonella spp. Associated with Chronic Periodontitis. Int J Curr Microbiol App Sci. 2015; 4(3): 150–160. Reference Source\n\nArif N, Sheehy EC, Do T, et al.: Diversity of Veillonella spp. from sound and carious site in children. J Dent Res. 2008; 87(3): 278–282. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "48110", "date": "20 May 2019", "name": "Juni Handajani", "expertise": [ "Reviewer Expertise Area of my expertise are oral biology and immunology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study is good for identification and phylogenetic analysis of oral Veillonella species isolated from saliva samples of children in Japan. The aim of this study was to analyze the composition and proportion of oral Veillonella species in the saliva of Japanese children compared to previous studies in Thailand.\n\nThe manuscript is certainly well written but I have some major concerns on the data analysis:\nThe number of samples in this study was fewer (15 children) compared to the number of samples in the previous study from Thailand (107 children). It is suggested that the analysis uses proportions so that it can describe the oral Veillonella species according to the number of samples.\n\nTo find out the comparison of the results of this study with the results of a previous study from Thailand, a correlation analysis is suggested.\n\nIt is also necessary to add the results of a correlation analysis between the results of Veillonella's oral identification and oral health status.\n\nMinor comment:\nSome references used are older than 10 years. I suggest to use current references from at least the last 10 years.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4665", "date": "20 Jun 2019", "name": "Ariadna Djais", "role": "Author Response", "response": "This study is good for identification and phylogenetic analysis of oral Veillonella species isolated from saliva samples of children in Japan. The aim of this study was to analyze the composition and proportion of oral Veillonella species in the saliva of Japanese children compared to previous studies in Thailand.Response: Thank you for your valuable comment. The aim of this study was focus to analyze the composition and proportion of oral Veillonella species in the saliva of Japanese children. Furthermore, we compared the results with some reports such as in Thailand. The revise version has been submitted.The manuscript is certainly well written but I have some major concerns on the data analysis: The number of samples in this study was fewer (15 children) compared to the number of samples in the previous study from Thailand (107 children). It is suggested that the analysis uses proportions so that it can describe the oral Veillonella species according to the number of samples. Response: These children visited to the Dental Hospital, Health Sciences University of Hokkaido for dental examination, over a period between 2016 and 2017. Based on the evaluation by the Simplified Oral Hygiene Index, they were divided into three groups, good, moderate and poor. But, among many children, only five children had poor oral hygiene status. Therefore, we used these five children as poor group. And we have selected five children from good and moderate groups, without distinction. As the results, the total subject : 15 subjects. To find out the comparison of the results of this study with the results of a previous study from Thailand, a correlation analysis is suggested. It is also necessary to add the results of a correlation analysis between the results of Veillonella's oral identification and oral health status. Response: Thank you for another good point at No. 2 and No.3. The research of oral Veillonella were very limited. The analysis was use in this study : descriptive analysis with mean value to determine the distribution and proportion of oral Veillonella, then we compare the characteristic of oral veillonella distribution with the previous study.Minor comment: Some references used are older than 10 years. I suggest to use current references from at least the last 10 years. Response: As we mentioned before, that the research of oral Veillonella were very limited. Therefore, some of references were used older that 10 years.Thank you for taking the time and energy to help us improve this manuscript. We hope that you will find these revision rise to your expectation." } ] }, { "id": "48113", "date": "23 May 2019", "name": "Takuichi Sato", "expertise": [ "Reviewer Expertise Oral Microbiology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nSuggestions:\n\nAbstract and Introduction: I think that the manuscript, in the Abstract and Introduction, needs to express rationales of this study in more detail, such as, why did the authors want to know the detection rate and distribution of oral Veillonella species in saliva of Japanese children? Why did the authors want to compare the results with those of Thai children?\n\nFigure 1 and Tables 1-3: I think that the data in Figure 1 and Tables 1-3 are overlapping, and Tables 1-3 could be deleted from the manuscript, because the authors mainly stated the total and the mean (proportions) of the isolates of Veillonella species in the Results and Discussion.\n\nMinor: The authors should check the meaning of the phrase \"although\":\nP. 7, Discussion: in the 6th paragraph L. 1-2: \"although 179 strains were isolated members of the genus Veillonella from saliva of 15 Japanese children, …\".\nP. 7, Discussion: in the 6th paragraph L. 6-7: \"although in this study 1,609 Veillonella strains were isolated from 107 Thai children\".\nP. 8, Discussion: in the 8th paragraph L. 7-8: \"although 10 unknown Veillonella strains analysed in this study formed three clusters distinct from V. dispar, …\".\n\nTypographical errors; P. 4, Results: Species identification L. 6: \"numberof\" should read \"number of\". (Put a space in between.)\nP. 7, Discussion: in the 5th paragraph L. 4: \"conforms\" should read \"conformed\".\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4666", "date": "20 Jun 2019", "name": "Ariadna Djais", "role": "Author Response", "response": "Abstract and Introduction:I think that the manuscript, in the Abstract and Introduction, needs to express rationales of this study in more detail, such as, why did the authors want to know the detection rate and distribution of oral Veillonella species in saliva of Japanese children? Why did the authors want to compare the results with those of Thai children?Response: Thank you very much for this direction. We have changed the wording of the relevant in the abstract to express the reason why did we want to know the detection rate and distribution of oral Veillonella species in Japan. The reasons are the study regarding the identification and distribution of oral Veillonella are limited; also the oral Veillonella community may affected by the differences in geographical location, age, diet, lifestyle, socio-economic status and oral hygiene status.Figure 1 and Tables 1-3:I think that the data in Figure 1 and Tables 1-3 are overlapping, and Tables 1-3 could be deleted from the manuscript, because the authors mainly stated the total and the mean (proportions) of the isolates of Veillonella species in the Results and Discussion.Response: Thanks for raising this important point. However, we consider to state the Table 1-3 as Ratio of the number of isolates of each species from each subject. Figure 1 was indicated to express the Total isolated number of each Veillonella species from each group of oral hygiene. Minor:The authors should check the meaning of the phrase \"although\":P. 7, Discussion: in the 6th paragraphL. 1-2: \"although 179 strains were isolated members of the genus Veillonella from saliva of 15 Japanese children, …\".Response: the word of “although” has been removedP. 7, Discussion: in the 6th paragraphL. 6-7: \"although in this study 1,609 Veillonella strains were isolated from 107 Thai children\".Response: the word of “although” has been removedP. 8, Discussion: in the 8th paragraphL. 7-8: \"although 10 unknown Veillonella strains analysed in this study formed three clusters distinct from V. dispar, …\".Response: the word of “although” has been removedTypographical errors;P. 4, Results: Species identificationL. 6: \"numberof\" should read \"number of\". (Put a space in between.)Response: the word of “numberof” has been revised to “number of”P. 7, Discussion: in the 5th paragraphL. 4: \"conforms\" should read \"conformed\".Response: the word of “conforms” has been revised to “conformed”Thank you for taking the time and energy to help us improve this manuscript. We hope that you will find these revision rise to your expectation." } ] } ]	1	https://f1000research.com/articles/8-616
https://f1000research.com/articles/7-1374/v1	31 Aug 18	{ "type": "Software Tool Article", "title": "epiflows: an R package for risk assessment of travel-related spread of disease", "authors": [ "Paula Moraga", "Ilaria Dorigatti", "Zhian N. Kamvar", "Pawel Piatkowski", "Salla E. Toikkanen", "VP Nagraj", "Christl A. Donnelly", "Thibaut Jombart", "Ilaria Dorigatti", "Zhian N. Kamvar", "Pawel Piatkowski", "Salla E. Toikkanen", "VP Nagraj", "Christl A. Donnelly", "Thibaut Jombart" ], "abstract": "As international travel increases worldwide, new surveillance tools are needed to help identify locations where diseases are most likely to be spread and prevention measures need to be implemented. In this paper we present epiflows, an R package for risk assessment of travel-related spread of disease. epiflows produces estimates of the expected number of symptomatic and/or asymptomatic infections that could be introduced to other locations from the source of infection. Estimates (average and confidence intervals) of the number of infections introduced elsewhere are obtained by integrating data on the cumulative number of cases reported, population movement, length of stay and information on the distributions of the incubation and infectious periods of the disease. The package also provides tools for geocoding and visualization. We illustrate the use of epiflows by assessing the risk of travel-related spread of yellow fever cases in Southeast Brazil in December 2016 to May 2017.", "keywords": [ "disease surveillance", "outbreaks", "epidemics", "infectious", "R", "RECON" ], "content": "Introduction\n\nInfectious disease outbreaks cause significant suffering and mortality in the affected populations, and damage the health, social and economic well-being of the families affected by diseases as well as producing significant economic costs for local and national governments. As we have seen with Ebola and SARS, disease outbreaks can spread beyond national borders1. Travelers can acquire a disease while staying in a foreign country, and then seed new outbreaks in their home country after their return. As international travel increases worldwide, new surveillance tools are needed to help identify locations where diseases are most likely to be spread and prevention measures need to be implemented. This is essential to limit the global spread of local outbreaks.\n\nRecently, Dorigatti et al.2 developed a method to assess the risk of travel-related international spread of disease by integrating epidemiological and travel (by air, land and water) volume. The model developed by Dorigatti et al.2 estimates the expected number of infections introduced elsewhere by taking into account population flows, lengths of stay, as well as the variability of the disease incubation and infectious periods. The method was applied to quantify the risk of spread of a recent outbreak of yellow fever in Southeast Brazil in December 2016 to May 2017, and was able to identify the countries that could have received travel-related disease cases capable of seeding local transmission.\n\nIn this paper we present epiflows, an R package that implements the method presented by Dorigatti et al.2 for risk assessment of travel-related spread of disease. Using data on population movement between the location that is source of the infection and other locations, lengths of stay, as well as information about the disease incubation and infectious period distributions, the package allows the estimation of the number of (symptomatic and/or asymptomatic) infections that could be spread to other locations together with uncertainty measures. The package also provides tools for geocoding and visualization of population flows.\n\nThe remainder of the paper is organized as follows. First, we briefly describe the modelling framework that is implemented in the epiflows package. Second, we introduce the main components of epiflows including instructions for installation and main functions. Third, we illustrate the use of the package via the assessment of the risk of travel-related spread of yellow fever cases due to population flows between Southeast Brazil and other countries in December 2016 to May 2017. Specifically, we discuss the data required and show how to perform the statistical analyses, how to interpret the results, and the visualization options. Finally, the conclusions are presented.\n\n\nModel\n\nIn this Section we explain the modelling framework presented in 2 for estimating the expected number of infections departing from one infectious location during the incubation or infectious periods. These cases comprise exportations and importations. Exportations refer to the infected residents of the infectious location (i.e. location with sustained disease transmission) that travel to other locations. Importations (also referred to as returning travelers) are people that are infected during a temporary stay in the infectious location and then return to their home location. The following Sections describe how to model exportations and importations to produce the total number of expected cases that could be spread to other locations together with uncertainty measures.\n\nLet CS,W denote the cumulative number of infections in location S in time window W. Here, W denotes the temporal window between the first and the last disease case in location S. Note that Dorigatti et al.2 calculated CS,W by multiplying the number of confirmed and reported yellow fever cases by 10 to account for underreporting of asymptomatic and mild yellow fever cases.\n\nLet popS be the resident population of the infectious location S, and TS,DW the number of residents of location S travelling to location D in time window W. The per capita probability that a resident from the infectious location travelled to other location D during the time window W is given by\n\n\n\nThe method assumes that the incubation period TE and the infectious period TI follow specific probability distributions. Using these, we can calculate the probability pi that an infection incubated or is infectious in time window W as\n\n\n\nFinally, the number of residents of the infectious location S that are infected and travel abroad during their incubation or infectious period during the time window W can be calculated as\n\nES,D = CS,W × pD × pi.\n\nThat is, ES,D is a product of the cumulative number of infections in location S in time window W, the per capita probability that a resident of S travels to location D, and the probability that an infection incubated or is infectious in time window W.\n\nNote here that if travel data are expressed annually (TS,DA) instead of in the time window W, travel data in the time window can be obtained as TS,DW = (TS,DA × W)/365.\n\nLet TO,SW be the number of travelers visiting location S from location O in time window W, and let LO denote the average length of stay. The per capita risk of infection of travelers visiting location S during their stay can be calculated as\n\n\n\nThe probability of returning to the home location while incubating or infectious is given by\n\n\n\nFinally, the expected number of travelers infected during their stay in the infectious location and returning to their home location before the end of the infectious period can be calculated as the product of the number of travelers, the per capita risk of infection and the probability of returning home while incubating or infectious,\n\n\n\nNote that, similarly to exportations, if travel data are expressed annually (TO,SA) instead of in the time window W, travel data in the time window can be obtained as TO,SW = (TO,SA × W)/365.\n\nFinally, the expected number of infections departing from the infectious location S to location O during the incubation or infectious periods can be computed as the sum of the number of infected residents of S travelling during their incubation or infectious periods, and the travelers from abroad that are infected during their stay in S and return to their origin location before the end of the infectious period. That is,\n\n\n\nAverage estimates and the relative uncertainty are calculated by taking into account the variability of the incubation and infectious periods. Specifically, the method samples a large number of times from the incubation and infectious distributions, which produces a full distribution for pi (the probability that a disease case is incubated or infectious in the time window considered) and pl (the probability of returning to the home location while incubating or infectious). This, in turn, creates variability in exportations ES,O and importations IS,O, and finally in the total number of infections introduced in location O, TS,O.\n\n\nMethods\n\nThe R package epiflows [17] is hosted in the Comprehensive R Archive Network (CRAN) which is the main repository for R packages: http://CRAN.R-project.org/package=epiflows. Users can install epiflows in R by executing the following code:\n\n\n\nThere is also a development version from GitHub which can be accessed at https://github.com/reconhub/epiflows. This version of the package may contain new features which are not incorporated in the version on CRAN yet but may be useful for some users. GitHub also includes issue tracking where users can note problems or suggestions for improvements. This development version from GitHub can be installed by using the install_github() function from the R package devtools3:\n\n\n\nWhen installing epiflows, other R packages which epiflows depends on are also automatically installed. These packages include sp4 for manipulating spatial objects; geosphere5 for calculating distances between locations; and leaflet6 for visualization.\n\nThe main function of the package is estimate_risk_spread() which calculates the mean and 95% confidence intervals of the number of cases spread to different locations from an infectious location. It is also possible to use this function to produce a data frame with all simulations (not just the mean and 95% confidence intervals that is computed from the simulations). This permits the user to aggregate the estimates and calculate confidence intervals with different levels using single simulations. To execute this function the following information is needed:\n\npopulation of the infectious location,\n\nnumber of infections in the infectious location, and the first and last dates of reported cases,\n\nnumber of travelers between the infectious location and other locations,\n\naverage length of stay of travelers from other locations visiting the infectious location,\n\ndistributions of the incubation and infectious periods,\n\nnumber of simulations to be drawn from the incubation and infectious period distributions,\n\nlogical value indicating whether the returned object should be a data frame with all simulations, or a data frame with the mean and lower and upper limits of a 95% confidence interval of the number of infections spread to each location.\n\nOther useful functions are plot() which produces visualizations of population flows between locations, and add_coordinates() which finds the coordinates of the locations.\n\n\nUse cases\n\nIn this Section we provide an example on how to use epiflows to calculate the number of yellow fever cases spreading from south-east Brazil to other countries due to human movement. We show how to define the arguments of the estimate_risk_spread() function, interpret the results, and make visualizations with the population flows.\n\nWe use the data YF_Brazil which is contained in the epiflows package as data(YF_Brazil). YF_Brazil is a list containing the population size, the assumed number of yellow fever infections, dates of first and last case reporting, number of travelers and length of stay for the states of Espirito Santo, Minas Gerais, Rio de Janeiro, Sao Paulo, and for the whole region of Southeast Brazil (which comprises the four states of Espirito Santo, Minas Gerais, Rio de Janeiro and Sao Paulo) in the period December 2016 to May 2017 [15], [16].\n\nFollowing Dorigatti et al.2, the total number of yellow fever infections in each of the Brazilian states was calculated by multiplying the cumulative number of confirmed yellow fever cases reported in 7 by 10 to account for underreporting of asymptomatic and mild yellow fever cases. The dates of first and last case reported in each state were derived as described by Dorigatti et al.2. Population data were obtained from the Brazilian Institute of Geography and Statistics website8. YF_Brazil also contains the number of travelers in the specified time window between the states of Espirito Santo, Minas Gerais, Rio de Janeiro, Sao Paulo (and the whole Southeast Brazilian region) and other countries. These estimates were obtained from World Tourism Organization data on the volume of air, land and water border crossings for Brazil for the year 20159, having assumed that travelers were distributed across the Brazilian states according to the relative population density and having accounted for information on the monthly distribution of tourism and on the average duration of stay of international visitors to Brazil10, as detailed in 2.\n\nYF_Brazil$states is a data frame that contains, for each Brazilian state considered in our example, the code (location_code), the population (location_population), the number of assumed infections in the time window (num_cases_time_window), and the dates of the first and last case reported (first_date_cases and last_date_cases, respectively).\n\n\n\n\n\n\n\nYF_Brazil$T_D represents TS,DW and contains the number of travelers from each of the Brazilian states to other countries. YF_Brazil$T_O represents TO,SW and contains the number of travelers from other countries to each of the Brazilian states.\n\n\n\n\n\nFinally, YF_Brazil$length_of_stay is a vector with the lengths of stay (in days) of the travelers visiting Brazil from other countries.\n\n\n\nTo aid in data organization between flows and metadata, we have implemented the “epiflows” object. This inherits the “epicontacts” object from the epicontacts package11, storing three elements:\n\n1. flows — a data frame defining the number of cases flowing from one location to another\n\n2. locations — a data frame listing the locations present in flows and relevant metadata.\n\n3. vars — a dictionary mapping column names in locations to known global variables defined in global_vars(). These global variables are used as default values in estimate_risk_spread().\n\nAn epiflows object can be created with the make_epiflows() function by providing a data frame flows with the number of travelers between locations, a data frame locations with information about the locations, and the names of the columns of data frame locations indicating the name of each variable.\n\nIn the data frame flows each row represents the number of travelers from one location to the next. flows has at least three columns: columns from and to indicating where the flow starts and ends, respectively, and column n indicating the number of travelers that are in the flow. We can create a data frame YF_flows with the population flows of the Brazil data as follows.\n\n\n\n\n\nIn data frame locations each row represents a location, and columns specify useful information about the locations such as ID, population, number of cases, dates and length of stay. locations must contain at least one column specifying the location ID used in the flows data frame. We can create the data frame YF_locations by combining the data frame YF_Brazil$states containing the Brazil states information, and the vector YF_Brazil$length_of_stay containing the lengths of stay.\n\n\n\n\n\nThen, we can create an epiflows object called Brazil_epiflows as follows.\n\n\n\nThe arguments that need to be specified in estimate_risk_spread() to calculate the cases or infections introduced in other countries are as follows. The first argument is an epiflows object containing the number of travelers between locations, the population size, the number of cases, and the first and last dates of reporting in the infectious location, and the average length of stay in days of travelers from other locations visiting the infectious location.\n\nThe second argument of estimate_risk_spread() is location_code which is a character string denoting the infectious location code. We also need to specify the incubation and infectious period distributions. Specifically, we need to provide functions with a single argument n that generate n random incubation and infectious periods. In this example, we assume that the incubation period TE is log-normally distributed with mean 4.6 days and variance 2.7 days, and that the infectious period TI is normally distributed with mean 4.5 days and variance 0.6 days. We can define functions incubation() and infectious() as\n\n\n\nArgument num_sim is the number of simulations to be drawn from the incubation and infectious period distributions. It is recommended to use at least 1,000 simulations. The last argument of estimate_risk_spread() is return_all_simulations. This is a logical value indicating whether the returned object should be a data frame with all simulations (return_all_simulations = TRUE), or a data frame with the mean and lower and upper limits of a 95% confidence interval of the number of infections spread to each location (return_all_simulations = FALSE).\n\nOnce we have constructed the objects needed to call estimate_risk_spread() we can execute the function and obtain the estimated mean number of cases spread to each country and the 95% confidence intervals. The code to calculate the cases spread from Espirito Santo is the following:\n\n\n\nThe results returned by estimate_risk_spread() are stored in the res object. This is a data frame with the columns mean_cases indicating the mean number of cases spread to each location, and lower_limit_95CI and upper_limit_95CI indicating the lower and upper limits of 95% confidence intervals. The result object is shown below.\n\n\n\n\n\nWe can plot the results with ggplot() as follows (Figure 1).\n\n\n\nNote that if we set return_all_simulations equal to TRUE, the result object res will be a data frame with all simulations.\n\n\n\n\n\n\n\nUsing res, we can calculate the mean and 95% confidence intervals as follows.\n\n\n\n\n\nWe can visualize flows of people travelling between locations using plot() and passing as first parameter an epiflows object containing the population flows, and as second parameter the type of plot we wish to produce. Population flows can be displayed on an interactive map, as a network or as a grid between origins and destinations as described in the following sections.\n\nWe can visualize population flows on an interactive map using plot() with the parameter type = \"map\". For this option to work, the epiflows object needs to include the longitude and latitude of the locations in decimal degree format. If coordinates are known, they can be added to the epiflows object using the add_coordinates() function from the epiflows package. In our example, the longitude and latitude data are in the data frame YF_coordinates.\n\n\n\nThey can be added as follows.\n\n\n\nIf coordinates are unknown, we may resort to one of the freely available tools for geocoding. For example, we can use the geocode() function from the ggmap package12. This function finds the latitude and longitude of a given location using either the Data Science Toolkit or Google Maps. We can also use add_coordinates() which uses geocode() to find the coordinates and directly add them to the epiflows object as follows.\n\n\n\nOnce we have assigned coordinates to the epiflows object, we can use plot() with type = \"map\" to visualize the population flows between locations in an interactive map (Figure 2).\n\n\n\nThe produced map can be zoomed and permits an easy examination of flows. plot() uses the gcIntermediate() function from the geosphere package5 to obtain the great circle arcs between locations, and then uses the leaflet package6 to create an interactive map with the connection lines. The connection lines are coloured according to flow volume, and as the mouse passes over the lines, lines highlight and information about connections is shown. We can also include parameters to specify a title, the center of the map or a color palette. An interactive version of this visualization is shown here: https://www.repidemicsconsortium.org/epiflows/articles/introduction.html#introduction-epiflows-map.\n\nPopulation flows can also be displayed as a dynamic network using plot() with type = \"network\" (Figure 3).\n\n\n\nThis option uses the package visNetwork13 to show the locations as nodes of a network and connections between them representing population flows. This plot is interactive and it is possible to highlight a given location and examine its population flows, as well as its population, number of cases, dates and length of stay. This type of plot can be used when coordinates of locations are missing. An interactive version of this plot can be viewed here: https://www.repidemicsconsortium.org/epiflows/articles/introduction.html#introduction-epiflows-vis.\n\nFinally, population flows can also be shown as a grid between locations with the option type = \"grid\" (Figure 4).\n\n\n\nThis plot shows flows between locations as points by positioning origins and destination in y and x axes, respectively. When using this option, additional arguments can be passed to set the size, color or shape of the points as in function geom_point() of package ggplot214. As the network plot, the grid plot can be used when coordinates of locations are missing.\n\n\nSummary\n\nIn this article we have presented the epiflows package for risk assessment of travel-related spread of disease. This package allows the estimation of the expected number of infections that could be introduced to other locations from the source of infection by integrating data on the number of cases reported, population movement, length of stay and information on the distributions of the incubation and infectious periods of the disease. The package also provides tools for geocoding and visualization which facilitate the interpretation of the results.\n\nFirst, we presented how to estimate exportations, importations and total number of infections using the modelling framework introduced by Dorigatti et al.2. Then, we demonstrated the use of the package by assessing the risk of travel-related spread of yellow fever cases in Southeast Brazil in December 2016 to May 2017. Specifically, we have shown how to construct an epiflows object containing population flows and information about locations, and how to use the function estimate_risk_spread() to obtain the average and confidence intervals of the estimated number of infections introduced elsewhere. Finally, we have shown how to visualize the results and produce maps of the population flows.\n\nInternational travel has an important role in the spread of infectious diseases across national borders. We think the epiflows package represents a useful tool for disease surveillance that can help public health officials identify locations where diseases are most likely to spread and prevention measures are most needed.\n\n\nData availability\n\nDataset 1. Arrivals of non-resident tourists at Brazilian national borders by country of residence. Annual volumes of air, land and water border crossings for Brazil relative to inbound tourism from years 2011 to 2015 obtained from the World Tourism Organisation. https://doi.org/10.5256/f1000research.16032.d21576315.\n\nDataset 2. Trips abroad by Brazilian resident visitors to countries of destination. Annual volumes of air, land and water border crossings for Brazil relative to outbound tourism from years 2011 to 2015 obtained from the World Tourism Organisation. https://doi.org/10.5256/f1000research.16032.d21576516.\n\n\nSoftware availability\n\n1. Dedicated website for epiflows, including installation guidelines and documentation: https://www.repidemicsconsortium.org/epiflows\n\n2. Software available from: https://cran.r-project.org/package=epiflows\n\n3. Source code available from: https://github.com/reconhub/epiflows\n\n4. Archived source code at time of publication: http://doi.org/10.5281/zenodo.140180617.\n\n5. Software license: MIT License", "appendix": "Author contributions\n\n\n\nPM, ZNK, PP and TJ developed the R package.\n\nST and VPN contributed to the R package.\n\nID and CAD developed the methods.\n\nID contributed data.\n\nPM, ID and ZNK analysed the data.\n\nPM wrote the first draft of the manuscript.\n\nAll authors read and approved the final manuscript.\n\n\nGrant information\n\nID acknowledges research funding from the Imperial College Junior Research Fellowship. ID, CAD and TJ thank the UK Medical Research Council for Centre funding. TJ is funded by the National Institute for Health Research Health Protection Research Unit (NIHR HPRU) in Modelling Methodology at Imperial College London in partnership with Public Health England (PHE).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe would like to thank the World Tourism Organisation for the permission to make public use of the border crossing data compiled by UNTWO and purchased on9. We would also like to thank the RECON consortium for establishing the research platform where this collaborative work was implemented.\n\n\nReferences\n\nHeymann DL, Chen L, Takemi K, et al.: Global health security: the wider lessons from the west African Ebola virus disease epidemic. Lancet. 2015; 385(9980): 1884–1901. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDorigatti I, Hamlet A, Aguas R, et al.: International risk of yellow fever spread from the ongoing outbreak in Brazil, December 2016 to May 2017. Euro Surveill. 2017; 22(28): pii: 30572. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWickham H, Chang W: devtools: Tools to Make Developing R Packages Easier. R package version 1.13.3. 2017. Reference Source\n\nPebesma EJ, Bivand RS: Classes and methods for spatial data in R. R News. 2005; 5(2): 9–13. Reference Source\n\nHijmans RJ: geosphere: Spherical Trigonometry. R package version 1.5-5. 2016. Reference Source\n\nCheng J, Karambelkar B, Xie Y: leaflet: Create Interactive Web Maps with the JavaScript ’Leaflet’ Library. R package version 1.1.0. 2017. Reference Source\n\nBrasilia: Ministério da Saúde. Portuguese: Monitoramento dos casos e óbitos de febre amarela no Brasil, informe n. 43/2017. [Monitoring of the cases and deaths due to yellow fever in Brazil, update n. 43/2017], 2017. Reference Source\n\nRio de Janeiro: Instituto Brasileiro de Geografia e Estatística (IBGE): Estimativas populacionais para os municípios e para as Unidades da Federação brasileiros em 01.07.2016. [Population estimates for the municipalities and for the Brazilian Federal Units on 1 July 2016.], 2016. Reference Source\n\nMadrid: UN World Tourism Organization (UNWTO): Yearbook of tourism statistics dataset. 2016. Reference Source\n\nMinistério do Turismo: Estudo da Demanda Turística Internacional 2015. Study of the international tourist demand 2015, 2017. Reference Source\n\nNagraj VP, Randhawa N, Campbell F, et al.: epicontacts: Handling, visualisation and analysis of epidemiological contacts [version 1; referees: 1 approved, 1 approved with reservations]. F1000Research. 2018; 7(566). Publisher Full Text\n\nKahle D, Wickham H: ggmap: Spatial visualization with ggplot2. R J. 2013; 5(1): 144–161. Reference Source\n\nAlmende BV, Thieurmel B, Robert T: visNetwork: Network Visualization using ’vis.js’ Library. R package version 2.0.4. 2018. Reference Source\n\nWickham H: ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York, 2009; ISBN 978-0-387-98140-6. Reference Source\n\nMoraga P, Dorigatti I, Kamvar ZN, et al.: Dataset 1 in: epiflows: an R package for risk assessment of travel-related spread of disease. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16032.d215763\n\nMoraga P, Dorigatti I, Kamvar ZN, et al.: Dataset 2 in: epiflows: an R package for risk assessment of travel-related spread of disease. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16032.d215765\n\nKamvar ZN, Piątkowski P, Jombart T, et al.: reconhub/epiflows: Version 0.2.1: First zenodo release (Version v0.2.1). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1401806" }	[ { "id": "38778", "date": "09 Oct 2018", "name": "Noam Ross", "expertise": [ "Reviewer Expertise Disease ecology", "disease modeling", "R programming an package design." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors present and provide a tutorial to epiflows, an R package for calculating the risk of travel-related disease export from an epidemic area. It is a useful implementation of an algorithm, with associated visualization tools. It is technically sound though the scaffolding around the core algorithm is somewhat over-engineered.\n\nBoth the package design and paper description imply this package is designed for rapid risk assessment. My comments are primarily in regard to the clarity of the description and the usability of the package API in this context.\nThe `global_vars()` function is a thin wrapper around R's `options()` mechanism that obfuscates what is actually happening, and the name itself is somewhat confusing. (There are many different uses and abuses of global variables in R). It is difficult to see how this function improves over simply telling the user that default variables are defined by `epiflows.vars` in R's options mechanism (epiflows.varnames might be clearer).\n\nWhile it is mentioned that the `epiflows` object inherits from `epicontacts`, it is not clear what this means and how it is relevant to the user. Given the epiflows object has different contents than the `epicontacts` object, it should be explained. I see in the package vignette that subsetting methods are inherited, but given that the object contents are different, this should be demonstrated in the paper. Otherwise it is hard to see what the advantage of this object is at all, over simply passing data frames to the algorithm function.\n\nThe paper (as well as the code demonstration and vignette in the package), spends considerable time on the conversion of the partially processed data in `YF_Brazil` to data appropriately structured to be used in the `make_epiflows()` and `estimate_risk_spread()` function. This is confusing and not very useful - most users will not have data in exactly the format of `YF_Brazil`, and it distracts from the description of the core package functionality. It would be far clearer to introduce and demonstrate the functions first using data in its ready-to-analyze form. If the authors wish to provide an example of an actual full workflow, they should start with actual raw source data from the the supplementary data. A useful addition would also be to describe possible sources of the data needed.\n\nThe distributions of incubation and infectious periods is important and glossed over rather quickly here. First, on page 3, the text reads, \"The method assumes that the incubation period TE and the infectious period TI follow specific probability distributions.\" This is unclear, I think \"disease-specific\" would be clearer. Moreover, in the code tutorial, these distributions are simply assumed. It would make more sense to describe why such distributions are selected and the source of the parameters, e.g., \"For yellow fever we choose these distributions based on clinical literature describing the disease course as X to Y days of incubation and V to Z days infectious period (Citation 1, Citation 2). Lognormal or Gamma distributions are typically used for these distributions...\"\n\nIt also confusing that both individuals traveling and incubation times are shown as T in the mathematical notation.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Partly\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4776", "date": "02 Aug 2019", "name": "Paula Moraga", "role": "Author Response", "response": "Thank you for your thorough review and insightful comments. We address each comment separately below. 1. In response to the comment regarding the `global_vars()` function, we think this is a fair point. The main advantage of the use of `global_vars()` is giving the user a way to easily recover the `epiflows.vars` option variables in the case that they made an error in specifying new variables (global_vars(reset = TRUE)). 2. Thank you for bringing to our attention that the use of epicontacts needs an explanation. Because a flow of cases from one location to another can be thought of as a contact with a wider scope, the `epiflows` object inherits from the `epicontacts` object, where locations are stored in the \"linelist\" element and flows are stored in the \"contacts\" element (though the user does not need to interact with these elements by name). By building on the epicontacts object, we ensure that all the methods for sub-setting an object of class `epicontacts` also applies to `epiflows`, reducing the maintenance effort. We have clarified this in Section “The epiflows object” of the manuscript. 3. In order to present an example that is as clear as possible for readers, we have modified the yellow fever example and now we do not describe datasets that do not have ready-to-analyze form. Now we start by describing the YF_flows and YF_locations objects which can be directly passed to the make_epiflows() function to create the Brazil_epiflows object. 4. Thank you for pointing out the importance of the distributions of incubation and infectious periods. In Section \"Exportations\" we have replaced \"the method assumes that the incubation period T_E and the infectious period T_I follow specific probability distributions.\" by \"We assume that the incubation period (D_E) and the infectious period (D_I) are random variables, with associated probability distributions that are disease-specific.\" In Section \"Arguments of the estimate_risk_spread() function\", we have cited the papers we used to choose the incubation and infectious period distributions of the yellow fever example. We decided not to provide data for incubation and infectious periods of other pathogens since we expect users to have some knowledge of the life history of the disease they want to apply it to. We have also mentioned that users should consider the literature carefully before deciding on appropriate distributions, and we reference two review papers that may be useful. 5. Thank you for the notation review. Now we have changed the notation of the incubation period T_E by D_E, and infectious period T_I by D_I, where letter D stands for duration. Now letter T is only used for travelers." } ] }, { "id": "38978", "date": "18 Oct 2018", "name": "Jon Zelner", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this paper, the authors have presented and provided use-case examples for an R package that allows the estimation of the number of infections spread via travel, given information on the prevalence of disease in the sending location and the rate of flow between two locations.\n\nAlthough the paper builds on a published method, it should provide some additional detail about the method in the introductory sections. Specifically, in the section titled \"Exportations\" on page 3, the authors say that \"the method assumes that the incubation period T_E and the infectious period T_I follow specific probability distributions.\"\nMore explanation of the modeling assumptions will allow potential users of this package to make an informed decision about whether it will be useful to them without going back to the original manuscript in which the method was described. In addition, the reuse of \"T\" to indicate the rate of travel in the paragraph above and the incubation and infectious periods is confusing and should be changed.\nFrom reading the paper and looking at the github repository, it seems like it is completely up to the user to specify the distribution of the incubation and infectious periods. One potentially helpful addition to the package would be the addition of data with estimates of these quantities for different pathogens where available, along with citations/dois for the data source. This would make this more user-friendly and amenable to rapid response.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Partly\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly", "responses": [ { "c_id": "4777", "date": "02 Aug 2019", "name": "Paula Moraga", "role": "Author Response", "response": "Thank you for your helpful and insightful comments. Please find our responses below. 1. Thank you for pointing out the choice of incubation and infectious period distributions needs clarification. In Section \"Exportations\" we have mentioned we assume that the incubation period and the infectious period are random variables, with associated probability distributions that are disease-specific. We have also cited the papers we used to choice the incubation and infectious period distributions of our yellow fever example. 2. Thank you for the notation review. Now we have changed the notation of the incubation period T_E by D_E, and infectious period T_I by D_I, where letter D stands for duration. Now letter T is only used for travelers. 3. We agree on the importance of the choice distributions of incubation and infectious periods. In Section \"Arguments of the estimate_risk_spread() function\", we have added that we can define these distributions using random generation functions of distributions that are implemented in R. However, we decided not to provide data for incubation and infectious periods of other pathogens since we expect users to have some knowledge of the life history of the disease they want to apply it to. We have mentioned that users should consider the literature carefully before deciding on appropriate distributions, and we reference two review papers that may be useful." } ] } ]	1	https://f1000research.com/articles/7-1374
https://f1000research.com/articles/8-160/v1	06 Feb 19	{ "type": "Research Article", "title": "Screening is not associated with reduced incidence of gonorrhoea or chlamydia in men who have sex with men (MSM); an ecological study of 23 European countries", "authors": [ "Chris Kenyon" ], "abstract": "Background: Increasing rates of antimicrobial resistance has motivated a reassessment of if intensive screening for gonorrhoea and chlamydia is associated with a reduction in the prevalence of these infections in men who have sex with men (MSM). Methods: Spearman’s correlation was used to evaluate the country-level correlation between the intensity of self-reported sexual transmitted infection (STI) screening in MSM (both anal and urethral screening, taken from a large internet survey of MSM) and the incidence (taken from ECDC surveillance figures) and prevalence (taken from a literature review of studies estimating prevalence in MSM attending STI clinics) of gonorrhoea and chlamydia. Results: The intensity of both anal and genital screening was found to be positively associated with country level gonorrhoea incidence rates (rho 0.74; p=0.0004; rho=0.73; p=0.0004, respectively) and Ct incidence rates (rho 0.71; p=0.001; rho=0.78; p=0.0001, respectively). No associations were found between anal or genital screening intensity and Ng prevalence in clinic populations (Table 2). Conclusions: We found no evidence of a negative association between screening intensity and the prevalence of gonorrhoea or chlamydia in MSM. Randomized controlled trials are urgently required to evaluate if the high antimicrobial exposure resulting from intensive screening programmes is justified.", "keywords": [ "Gonorrhoea", "chlamydia", "MSM", "STI screening", "PrEP", "antimicrobial resistance" ], "content": "Introduction\n\nThere have been large increases in antimicrobial resistance in a number of sexual transmitted infections (STI) in the recent past. There are serious concerns that both Neisseria gonorrhoeae (Ng) and Mycoplasma genitalium may become untreatable in the not too distant future1,2. For both these bacteria as well as macrolide resistance in Treponema pallidum, AMR has frequently first emerged in populations with a combination of high antimicrobial consumption and dense sexual networks3,4. HIV preexposure prophylaxis (PrEP) cohorts have dense sexual networks and the intense screening STI typically practiced translates into high antimicrobial exposures5,6. Three-monthly, 3-site Ng/Chlamydia trachomatis (Ct) screening for example translates into macrolide exposures of around 4400 standard units/1000 population/year, which is many times higher than levels associated with the induction of macrolide resistance in a range of bacteria including T. pallidum and Ng7,8. These findings have led a number of authors to review the evidence to support Ng/Ct screening in men who have sex with men (MSM) PrEP populations.\n\nThe US Preventive Task Force, concluded that there is insufficient evidence to advocate for or against screening for Ng in men, including MSM9. In a systematic review conducted to inform these guidelines, the authors found no randomised, controlled trials or controlled observational studies that assessed the utility of NG screening in men10. In a systematic review of observational studies, we found no evidence that even the most intense Ng/Ct screening such as screening 100% of PrEP cohorts every 3 months was associated with a decline in the prevalence of these infections11. Others have argued that this lack of an effect was because the PrEP recipients were having sex with (and being reinfected by) people who were not being screened12. This generates the hypothesis that we test in this paper that populations where a high proportion of MSM are screened for Ng/Ct will have a lower prevalence of these infections than populations with less screening. We test this hypothesis in European countries because the intensity of STI screening varies widely here and data for screening and prevalence estimates were available.\n\n\nMethods\n\nSTI screening intensity. Country level STI screening prevalence were obtained from the European MSM Internet Survey (EMIS), which was an internet-based survey of over 160 000 MSM from 38 countries living in Europe13. The survey was conducted between June and August 2010. In the section where participants were asked about STI testing in the past 12 months, they were asked 3 questions that are relevant to Ng/Ct screening: Did you provide a urine sample for STI screening? Was urethral swab inserted into your penis for STI screening? Was a swab inserted into your anus for STI screening? EMIS combined the results from the first two questions into one variable reporting the proportion of respondents reporting ‘urethral STI screening’ – via either urine or urethral swab. The third question provided the proportion with ‘anal STI screening’. Typically, these urethral and anal samples are tested for Ng/Ct.\n\nNg/Ct prevalence/incidence.\n\n1. National Ng and Ct incidence estimates for men in 2010 were taken from European Centre for Disease Prevention and Control (ECDC) figures14. These incidence estimates are based on national surveillance systems. The ECDC does not provide incidence estimates separately for MSM and thus we used the estimates for all men. MSM do however constitute a high proportion of diagnoses in all men14.\n\n2. Systematic review of Ng/Ct prevalence in MSM\n\nNg/Ct prevalence estimates for MSM were taken from a published literature review of pharyngeal and anorectal Ng and Ct prevalence estimates in MSM (and other populations)15. All studies listed in PubMed reporting prevalence of extragenital Ng and Ct in MSM up to 1 December 2015 were included. A total of 53 studies were included of which 18 were from 6 European countries (Table 1). For the four European countries with more than one study we selected the study reporting prevalence estimates from 2010 or as soon after this year as possible. All selected studies were prevalence estimates established by Nucleic Acid Amplification Testing of MSM clients attending STI clinics.\n\nN. gonorrhoeae - Neisseria gonorrhoeae, C. trachomatis - Chlamydia trachomatis\n\nIn all analyses the correlation between screening intensity and Ng/Ct prevalence/incidence was tested using Spearman’s correlation. The statistical analyses were performed in STATA 13.\n\n\nResults\n\nThe proportion of respondents in each of the 23 countries reporting anal STI screening varied widely from 6.5 to 70.6% to (median 17.3%, IQR 11.8–47.1; Table 1). Likewise, there were large variations in the proportion reporting genital STI screening (range 37.0 to 94.0%, median 63.6 IQR 50.0–85.0%). There was a strong correlation between the proportions reporting anal and genital STI screening (rho=0.81; p<0.0001).\n\nFor 19 countries with data, the incidence of Ng for men in 2010 ranged between 1.2 and 42.2 cases per 100 000 men per year (median 7.2, IQR 3.6–18.3). There was an even wider distribution in estimated Ct incidence for the 18 countries with data (range 0 to 383.7, median 24.8, IQR 1.3–201).\n\nThere was less variance in the prevalence estimates of Ng and Ct in MSM (Table 1). Rectal Ng: median 5.5%, IQR 4.6–7; pharyngeal Ng: median 5.4%, IQR 3.9–6.5; urethral Ng: median 1.9%, IQR 1–3.4. Rectal Ct: median 7.3%, IQR 6.5–10.0; pharyngeal Ct: median 1.3, IQR 0.8–1.7; urethral Ct: median 3, IQR 2.5–5.3; Table 1.\n\nThe intensity of both anal and genital screening was found to be positively associated with country level Ng incidence rates (rho 0.74; p=0.0004; rho=0.73; p=0.0004, respectively) and Ct incidence rates (rho 0.71; p=0.001; rho=0.78; p=0.0001, respectively; Figure 1).\n\nCountry designations: AT, Austria; BE, Belgium; CZ, Czech Republic; DE, Germany; DK, Denmark; EL, Greece; ES, Spain; FI, Finland; FR, France; HR, Croatia; HU, Hungary; IE, Ireland; IT, Italy; LU, Luxembourg; LV, Latvia; NL, the Netherlands; PT, Portugal; SE, Sweden; SI, Slovenia; SK, Slovakia; UK, United Kingdom.\n\nNo associations were found between anal or genital screening intensity and Ng prevalence in clinic populations (Table 2).\n\nAll P-values were greater than 0.1.\n\n\nDiscussion\n\nA key reason for screening for Ng and Ct in MSM is to reduce the incidence and prevalence of these infections. In this analysis, we did not find evidence of a negative correlation between the intensity of STI screening in MSM and the incidence/prevalence of Ng/Ct. Instead, we found evidence of a positive association between the intensity of screening in MSM and the estimated incidence rate for men. This positive association may be explained by the fact the incidence estimates are influenced by the intensity of screening – countries with more intensive screening programmes would be expected to diagnose more asymptomatic Ng and Ct infections which lead to higher incidence estimates.\n\nTo deal with this bias and the fact that the ECDC Ng/Ct incidence estimates do not provide incidence estimates for MSM, we also evaluated the association in MSM attending STI clinics. Here we found no evidence of an association between screening intensity and prevalence.\n\nThese findings are open to a number of interpretations. Firstly, screening intensity may be negatively associated with Ng/Ct rates but we missed this association due to methodological issues. Our estimates of screening intensity were based on a single cross-sectional source. Although EMIS had a large sample size and the accuracy of its prevalence estimates for other variables has been validated in other studies13,16, these screening estimates may be inaccurate and may have changed over time. As noted above, the STI incidence estimates were for all men and were likely strongly influenced by practices such as screening intensity. The STI prevalence estimates in MSM were all taken from men attending STI clinics and thus are likely higher than general populations of MSM. The study design of each of the 6 studies contributing Ng and CT prevalence estimates differed somewhat further limiting the extent to which comparisons can be made between prevalence estimates derived from these studies. We could find no comparable data on the prevalence of Ng or Ct in general MSM populations.\n\nAlternatively, screening intensity as measured may not be associated with reduced Ng/Ct rates in MSM. Randomized controlled trials of the efficacy of screening for Ct in women on the prevalence of Ct have produced equivocal results17–20. Although no RCTs have been conducted in MSM, a systematic review of observational studies revealed that Ng/Ct screening, even when conducted at 3-sites every 3-months, was not associated with reductions in the prevalence of Ng or Ct11. If we consider Ng, numerous aspects of the way it circulates in contemporaneous populations of MSM may explain why screening has little or no effect on prevalence. Symptomatic disease is thought to typically occur soon (2–21 days) after infection and if symptoms do not develop the infection (particularly in the pharynx and rectum) tends to persist in a low abundance, low infectious state for up to 6 months21. Highly exposed individuals develop a type-specific immunity, but this immunity is largely ineffective in low exposure individuals21,22. As a result, the vast majority of Ng infections are asymptomatic and self-limiting in MSM PrEP populations21,23. Similar considerations apply to Ct. In the case of Ct there is however better evidence that treatment of Ct results in “arrested immunity” and thereby paradoxically increases the probability of reinfection24,25. If screening results in ‘arrested immunity’ it may paradoxically increase Ng/Ct prevalence/symptomatic disease. The sexual networks of PrEP recipients are very dense (up to a mean of 18 partners per 3 months26) and this is responsible for generating the high prevalences of Ng and Ct5. Removing individuals piecemeal from this network for screening and treating has no effect on the underlying determinant of high prevalence. As a result, the probability of reinfection and prevalence remaining high.\n\nMathematical models of Ng and Ct transmission in European countries like Belgium have thus found that the sexual network of MSM was so dense that current levels of Ng screening were having little to no effects on Ng prevalence27. In contrast, a modelling study from the United States found that 6-monthly screening of an expanded number of PrEP recipients could avert 40% of Ng and Ct infections28. This study did not however model pharyngeal transmission of Ng (which plays a major role in transmission) and did not model the impact of immunity or Ng’s ability to adapt to antibiotic pressure. These omissions may explain the discrepancy between its finding and that of the systematic review of observational studies which found that even 3-monthly screening was not associated with a decline in Ng or Ct prevalence11.\n\nBased on the findings of this study and those reviewed here we conclude that we cannot exclude the possibility that intense screening (at least 3-site, 3-monthly) may have a small to moderate influence on Ng/Ct prevalence in MSM. Randomized controlled trials are urgently required to test this hypothesis. In the interim, given the mounting evidence that Ng/Ct screening does not have a large effect on prevalence but does result in high levels of antimicrobial exposure, consideration should be given to reducing the intensity or stopping Ng/Ct screening in MSM in a phased and controlled manner that allows a detailed evaluation of the risks and benefits of screening.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.", "appendix": "Grant information\n\nThe author declared that no grants were involved in supporting this work.\n\n\nReferences\n\nEyre DW, Sanderson ND, Lord E, et al.: Gonorrhoea treatment failure caused by a Neisseria gonorrhoeae strain with combined ceftriaxone and high-level azithromycin resistance, England, February 2018. Euro Surveill. 2018; 23(27): 1800323. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBradshaw CS, Horner PJ, Jensen JS, et al.: Syndromic management of STIs and the threat of untreatable Mycoplasma genitalium. Lancet Infect Dis. 2018; 18(3): 251–2. PubMed Abstract \| Publisher Full Text\n\nLewis DA: The role of core groups in the emergence and dissemination of antimicrobial-resistant N gonorrhoeae. Sex Transm Infect. 2013; 89 Suppl 4: iv47–51. PubMed Abstract \| Publisher Full Text\n\nKenyon C: Prevalence of macrolide resistance in Treponema pallidum is associated with macrolide consumption. J Med Microbiol. 2018. PubMed Abstract \| Publisher Full Text\n\nKenyon C, Schwartz IS: Effects of Sexual Network Connectivity and Antimicrobial Drug Use on Antimicrobial Resistance in Neisseria gonorrhoeae. Emerg Infect Dis. 2018; 24(7): 1195–1203. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKenyon C: Risks of Antimicrobial Resistance in N. gonorrhoeae Associated with Intensive Screening Programs in Pre-Exposure Prophylaxis Programs. Clin Infect Dis. 2018; 67(1): 154–5. PubMed Abstract \| Publisher Full Text\n\nKenyon C: We need to consider collateral damage to resistomes when we decide how frequently to screen for chlamydia/gonorrhoea in preexposure prophylaxis cohorts. AIDS. 2019; 33(1): 155–7. PubMed Abstract\n\nKenyon C, Buyze J, Wi T: Antimicrobial Consumption and Susceptibility of Neisseria gonorrhoeae: A Global Ecological Analysis. Front Med (Lausanne). 2018; 5: 329. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeFevre MLU.S. Preventive Services Task Force: Screening for Chlamydia and gonorrhea: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med. 2014; 161(12): 902–10. PubMed Abstract \| Publisher Full Text\n\nZakher B, Cantor AG, Pappas M, et al.: Screening for gonorrhea and Chlamydia: a systematic review for the U.S. Preventive Services Task Force. Ann Intern Med. 2014; 161(12): 884–93. PubMed Abstract \| Publisher Full Text\n\nTsoumanis A, Hens N, Kenyon CR: Is Screening for Chlamydia and Gonorrhea in Men Who Have Sex With Men Associated With Reduction of the Prevalence of these Infections? A Systematic Review of Observational Studies. Sex Transm Dis. 2018; 45(9): 615–622. PubMed Abstract\n\nRidpath AD, Chesson H, Marcus JL, et al.: Screening Peter to Save Paul: The Population-Level Effects of Screening Men Who Have Sex With Men for Gonorrhea and Chlamydia. Sex Transm Dis. 2018; 45(9): 623–5. PubMed Abstract \| Free Full Text\n\nThe EMIS Network: The European MSM Internet Survey (EMIS) Community Report. Stockholm: European Centre for Disease Prevention and Control, 2013 Contract No.: 01/04/2014.\n\nEuropean Centre for Disease Prevention and Control: Sexually transmitted infections in Europe 1990–2010. Stockholm: ECDC. 2012. Reference Source\n\nChan PA, Robinette A, Montgomery M, et al.: Extragenital Infections Caused by Chlamydia trachomatis and Neisseria gonorrhoeae: A Review of the Literature. Infect Dis Obstet Gynecol. 2016; 2016: 5758387. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMarcus U, Hickson F, Weatherburn P, et al.: Estimating the size of the MSM populations for 38 European countries by calculating the survey-surveillance discrepancies (SSD) between self-reported new HIV diagnoses from the European MSM internet survey (EMIS) and surveillance-reported HIV diagnoses among MSM in 2009. BMC Public Health. 2013; 13: 919. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAndersen B, van Valkengoed I, Sokolowski I, et al.: Impact of intensified testing for urogenital Chlamydia trachomatis infections: a randomised study with 9-year follow-up. Sex Transm Infect. 2011; 87(2): 156–61. PubMed Abstract \| Publisher Full Text\n\nvan den Broek IV, van Bergen JE, Brouwers EE, et al.: Effectiveness of yearly, register based screening for chlamydia in the Netherlands: controlled trial with randomised stepped wedge implementation. BMJ. 2012; 345: e4316. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLow N, Redmond S, Uusküla A, et al.: Screening for genital chlamydia infection. Cochrane Database Syst Rev. 2016; (9): CD010866. PubMed Abstract \| Publisher Full Text\n\nHocking JS, Temple-Smith M, Guy R, et al.: Population effectiveness of opportunistic chlamydia testing in primary care in Australia: a cluster-randomised controlled trial. Lancet. 2018; 392(10156): 1413–22. PubMed Abstract \| Publisher Full Text\n\nHook EW, Handsfield H: Gonococcal infections in the adult. In: Holmes KK, editor. Sexually transmitted diseases. 3rd ed. New York: McGraw-Hill, Health Professions Division; 1999; xxiii, 1454, 118.\n\nPlummer FA, Simonsen JN, Chubb H, et al.: Epidemiologic evidence for the development of serovar-specific immunity after gonococcal infection. J Clin Invest. 1989; 83(5): 1472–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFairley CK, Hocking JS, Zhang L, et al.: Frequent Transmission of Gonorrhea in Men Who Have Sex with Men. Emerg Infect Dis. 2017; 23(1): 102–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGeisler WM, Lensing SY, Press CG, et al.: Spontaneous resolution of genital Chlamydia trachomatis infection in women and protection from reinfection. J Infect Dis. 2013; 207(12): 1850–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOmori R, Chemaitelly H, Althaus CL, et al.: Does infection with Chlamydia trachomatis induce long-lasting partial immunity? Insights from mathematical modelling. Sex Transm Infect. 2018; pii: sextrans-2018-053543. PubMed Abstract \| Publisher Full Text\n\nGrant RM, Lama JR, Anderson PL, et al.: Preexposure chemoprophylaxis for HIV prevention in men who have sex with men. N Engl J Med. 2010; 363(27): 2587–99. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBuyze J, Vanden Berghe W, Hens N, et al.: Current levels of gonorrhoea screening in MSM in Belgium may have little effect on prevalence: a modelling study. Epidemiol Infect. 2018; 146(3): 333–338. PubMed Abstract \| Publisher Full Text\n\nJenness SM, Weiss KM, Goodreau SM, et al.: Incidence of Gonorrhea and Chlamydia Following Human Immunodeficiency Virus Preexposure Prophylaxis Among Men Who Have Sex With Men: A Modeling Study. Clin Infect Dis. 2017; 65(5): 712–718. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "44908", "date": "19 Mar 2019", "name": "Hamish McManus", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a well written study aimed to measure correlation between country-level sexually transmissible infection (STI) screening intensity in MSM, and country-level incidence of Neisseria gonorrhoeae (NG) and Chlamydia trachomatis (CT) in MSM in European countries. The study sought to address a hypothesised association between risk of reinfection by unscreened cases and the intensity of screening in those populations. This hypothesis is consistent with no prior studies having detected negative correlation between screening and prevalence of STIs. While the study is ultimately unable to draw conclusions on the association between screening intensity and prevalence, its importance lies in potentially motivating discussion and further examination of this topic. One of the most important points it illustrates is that findings based on cross-sectional data such as these are open to a number of interpretations, and hence motivate more thorough analyses using longitudinal data.\n\nIn part the reason for the fairly limited conclusions and presentation of a range of interpretations of the findings lies with limitations of the ecological source data. For example, the source data used for incidence estimation should be interpreted cautiously. The ECDC estimates cited are, precisely, national-level notification rates for the general male population rather than incidence rates for the respective MSM population. This is problematic for several reason. Firstly, the association between increased screening intensity and notification rates has been established by previous studies and estimates of incidence based directly on notification rates need to be adjusted accordingly to reflect this.1,2 And, secondly, it is not made clear what effect the use of general male population ECDC estimates rather than MSM specific estimates has on results. Although authors suggest that the ECDC estimates are strongly weighted by MSM, this statement is not supported. While MSM are disproportionately represented in NG/CT notifications, it is still probable that the strength of correlation between screening in this population and general incidence rates could be reduced substantially given that MSM comprise a very low (<7%) proportion of the male population.\n\nIn part to address these limitations (“To deal with this bias and the fact that the ECDC NG/CT incidence estimates do not provide incidence estimates for MSM” [P5]), the correlation between screening and prevalence estimates in MSM from STI clinics was also evaluated. However, this analysis is relatively limited, and we are not sure that it addresses these concerns successfully. Specifically, the prevalence results do not improve the interpretability of the incidence results. Also, they are based on complete results from only 4 countries (2 of which are not included in the incidence comparison) which is likely to limit the levels of correlation capable of being determined.\n\nTo overcome the limitations of this ecological study, the author concludes that randomised controlled trials are urgently required. However, as this would require abandoning STI screening for some of the participants, such a trial would be ethically dubious and contrary to current clinical guidelines. Alternatively, longitudinal administrative data can be subject to retrospective cohort analysis. Using this technique, we were able to determine a true national increase in the incidence of NG in MSM but, after controlling for test frequency, this could be explained by increasing partner numbers and condomless anal sex.3\n\nHowever, the statistical analyses in Kenyon’s study are appropriate and robust. More complex methods may have been inappropriate for the broad ecological source data used. Appropriately, the conclusions drawn are careful and, in that regard, supported by the results. This study should proceed to being indexed for the reason that it motivates discussion and should motivate more rigorous research into these growing epidemics.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "53339", "date": "03 Sep 2019", "name": "Eline L Korenromp", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nGeneral Comments: This is a useful study, well designed, analyzed, and written, ending in a well-reasoned call for improving the evidence base for gonorrhea and chlamydia screening programs.\n\nThe discussion adds particular value, comparing and contrasting findings from empirical studies with predictions by modelling studies. This part could possibly be structured better (e.g. I found it confusing to see reference 11 discussed twice on page 6). The authors might add a general statement about the value of mathematical models to answer complex epidemiological questions such as this. For example \"This illustrates that mathematical models to assess health interventions may be wrong for reasons that are not always obvious, and underlines the importance of empirical evidence\".\n\nThe specific comments below may serve to improve readability and interpretation.\n\nMethods: Page 3, right column, top: ‘A total of 53 studies were included of which 18 were from 6 European countries’. I suppose the remaining 35 studies were from other European countries, or also non-European? Perhaps just drop ‘European’ from this sentence, so to avoid this confusion.\n\nWhile, according to reference 14, a large proportion of Gc/Ct diagnoses takes places in MSM, it is unlikely that this reflects actual incidence, especially with regard to Ct which is highly prevalent, often asymptomatically, in the heterosexual population (e.g. in adolescent girls). This may be worth mentioning (in Introduction or Discussion) – to put the importance of MSM in the overall epidemic in context.\n\nResults: Page 4: ‘No associations were found between anal or genital screening intensity and Ng prevalence in clinic populations’. You found the same for CT, right? Please add that.\n\nDiscussion: Page 4: The authors offer as a possible explanation for the observed positive (and not, the hoped negative) association that screening makes people susceptible to reinfection. To me a more basic explanation in this ecological analysis based on cross-sectional data, is screening programs are likely targeted to regions and populations with high prevalence. Reverse causality: the high prevalence is likely the cause rather than the effect of the screening program being there.\n\nPage 5: ‘To deal with this bias … we also evaluated the association in MSM attending STI clinics’: Not clear, the association between what and what? You mean, between national screening intensity and prevalence in MSM attending STI clinics? Please rephrase, and refer back to Table 2 here.\n\nPage 5: ‘As noted above, the STI incidence estimates were for all men and were likely strongly influenced by practices such as screening intensity’. A more general limitation of these incidence estimates based on national surveillance data is that NG and Ct incidence is hardly measurable, and case notifications are no good indication of underlying incidence of these infections, which are most often asymptomatic and not presenting to clinics. Besides variations in screening intensity, variations across countries in health care access and population awareness of STIs contribute to varying case notification rates, which do not reflect true variations in underlying incidence, and so may bias (or at least dilute the power) of the ecological analysis.\n\nPage 6: ‘… limiting the extent to which comparisons can be made between prevalence estimates’: Please consider to rephrase as ‘… limiting the extent to which correlations could be assessed between screening intensity and prevalence across these studies’\n\nPage 6: About the systematic review of observational studies, quoted as reference 11, could you summarize how this differs from your current study, in methodology, scope, populations covered and/or other aspect?\n\nPage 6: ‘These omissions may explain the discrepancy between its finding and that of the systematic review of…’ This sentence may be clearer if written as: ‘… between its prediction, and our and the earlier systematic review of observational studies’.\n\nConclusion: The first concluding sentence is a perhaps optimistic twist to a negative finding, which may surprise some readers (such as me). It does serve as a good introduction to the authors’ call for randomized trials to test the impact of screening in MSM. For readability, the authors may consider to a few word changes: ‘Based on the findings of this study and those reviewed we conclude that we can STILL not exclude...’.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-160
https://f1000research.com/articles/8-1623/v1	11 Sep 19	{ "type": "Research Article", "title": "Effects of self-experimentation during practical classes on student learning", "authors": [ "A.M. de Leão e Neves Eduardo", "E.C. Campos Caldas Rosa", "A. Fonseca Welker" ], "abstract": "Background: This article reports an experiment based on the measurement of the academic achievement of students submitted to self-experiments during practical classes compared with students who attended regular practical classes (control group) to assess whether this intervention could help assess the influence of self-experiments on student learning. Methods: This study involved 71 students in the second terms of the degree of Bachelor of Pharmaceutical Sciences, studying the Cell Biology unit. Experiments were conducted using the students themselves as subjects under conditions that mimic situations observed in daily life, fasting and ingestion of carbohydrates. The performance of students in health college student assessments and the degree of motivation in performing these experiments was assessed at the Anhanguera college, Brasilia, Brazil. In total, 33 students (46.5%) participated actively in the experiment and the rest only observed the experiment carried out by the first group. In Cell Biology, the students study cell types, cell components and their respective functions, during one term, with a total workload of 60 h. Results: In the test that preceded the experiment carried out in the present study, the grades of the students that participated actively and of those that only observed were not statistically different (p > 0.05). In the test applied after the experiment, both groups reached higher grades (p < 0.01), but the individuals that participated actively in the experiment obtained higher grades than those that only observed it (p < 0.05). Conclusions: The findings of this study indicate that active learning, in which the students use their own organism and blood in practical classes, can increase their grades in knowledge tests. The teaching tool showed in the present study is a positive alternative for university students in health sciences.", "keywords": [ "Self-experiment", "active-learning method", "motivation to study", "practical class" ], "content": "Introduction\n\nSeveral studies have shown that active learning is more attractive and generates more interaction than passive methods1–4. Among active-learning strategies, some studies5,6 point out that practices that mimic routine activities increase learning compared to practices without this characteristic. For example, the use of virtual patients by medical students can increase their degree of learning5, and the use of case-based learning increases deep learning, the involvement of students, and their interest in content6,7. Problem-based learning (PBL), which frequently involves simulations of real situations (e.g., diagnosis of a hypothetical patient with some symptoms), has been an important tool to improve knowledge retention8–10. Another active-learning strategy that may increase academic achievement is the use of self-monitoring and self-assessment methods11–14. For example, after answering questions of a knowledge test, when students are allowed to see their colleagues’ answers and revise their own answers, the number of correct answers increases15. Evidence indicates that practices that involve both routine activities and self-monitoring promote even greater attraction of students for classes16,17, as well as greater interaction during classes1. A previous study submitted university students to the ingestion of pizza and pasta and the measurement of biochemical parameters18. This method made it possible to cover important topics of metabolism in the final exam and to show how the metabolic machinery of the human organism reacts to different diets. In the period when this method was applied, the grades were higher than in previous periods. However, there was no control group, which hindered assessing the real efficiency of this method18. Other studies in which the students used their own body (self-experiments) did not investigate the influence of this practice on learning19. The few reports on self-experiments frequently involve activities that are not part of the students’ routine20.\n\nAs far as we know, there are no studies on the impact on learning of practical classes using the students’ own bodies in routine situations. A experiment based on the measurement of the academic achievement of students submitted to self-experiments during practical classes compared with students who attended regular practical classes (control) could help assess the influence of self-experiments on student learning. The present study shows a new active-learning method, which is more efficient than the methods commonly used in classrooms.\n\nThe present study aimed at analyzing the impact of self-experiments under conditions that mimic everyday situations – fasting and ingestion of carbohydrate-rich oral solution with glucose and fructose – on the achievement in exams of university students of health sciences and their degree of motivation and satisfaction in carrying out those experiments.\n\n\nMethods\n\nTo analyze the impact of self-experimentation on learning in health sciences, randomly selected students from one class of university students studying Pharmacy participated actively in one experiment, whereas the other part only followed the study as observers. The students actively participating in the research were randomly dived into three subgroups: groups 1 and 2 (those who ingested glucose and fructose oral solutions, respectively) and group 3 (the fasting group). The experiment allowed the students to study some of the content comprised in Cell Biology. It involved the collection of blood samples of the students at times that mimic everyday situations – fasting and ingestion of different carbohydrates. In addition, the effect of the experiment on the degree of motivation and satisfaction of students was assessed. According to a sample calculation performed using GPower 3.1 software, the group should contain a total of 71 students of the second term the Bachelor of Pharmaceutical Sciences (33 active and 38 observers 38) to achieve a statistical power of 60% to achieve the objectives. A total of 33 students (46.5%) were randomly assigned to participate actively in the experiment and the rest observed the experiment carried out by the first group. In Cell Biology, the students study cell types, cell components and their respective functions, during one term, with a total workload of 60 h. The assessment of the students’ knowledge levels occurred through two written, closed-notes, and closed-book tests and the application of questionnaires on the topics studied in the classroom. The tests were the same for all groups. The study took place in March-April 2015.\n\nAfter eight weeks of classes, the teacher applied the first individual test that assessed the knowledge level of the students about cell biology, which included lipids, carbohydrates, and cell organelles. The test results were revealed in the next class, one week later. After the first written test, the teacher (A.M.L.N.E.) told the students about the experiment that she intended to carry out and all agreed to participate. An anthropometric questionnaire was applied to the students, after they agreed to participate in the project and provided written informed consent. Then, a biochemical experiment was performed by the students. All students that participated actively in the experiment were asked to fast overnight for 12 h. The first capillary blood sample was collected in the morning after 12 h of fasting, at the beginning of class, by some trained students of the same class together with the researcher responsible for the project. The students that took part in the experiments had two other capillary blood samples collected at 30 and 120 min and were randomly assigned to one of the following conditions: fasting (fasting group), ingestion of 1 g/kg of glucose solution (glucose group, 17 students), and ingestion of 1 g/kg fructose solution (fructose group, 16 students). The fructose and glucose solutions were prepared with 5 ml/kg of water. The control group only observed sample collection and the experiment, i.e., observed all procedures, but did not donate blood or ingest any solution. The experiment was carried out in two class days separated by a week. In addition to taking part in or only observing the experiments, all students studied the content of the subject cell biology under the same teaching methodology: audiovisual aided lectures, in which the teacher allowed and stimulated student participation with questions and observations. The subtopics taught were: lipids, carbohydrates, proteins, enzymes, and cell metabolism, all correlated with the central topic of the subject. After the practical/experimental part of the course, all students answered a questionnaire, available as Extended data21, in the classroom, with questions about motivation and the importance of the practical class to learning. After carrying out the experiment and answering the questionnaire in the classroom, all students did a second individual test with the contents of the subject Cell Biology.\n\nThis research project was approved by the Research Ethics Committee registered through the Research Ethics Committees (CEPs, acronym in Portuguese) and the National Research Ethics Commission (CONEP, acronym in Portuguese) also known as the CEP/CONEP system (approval number 19059513.3.0000.5372 and expert review nº 434.935). Written informed consent from all subjects involved was obtained for participation in the study.\n\nTo test for the homogeneity of qualitative variables between the “control/observer” and “active” groups, we used the Fisher exact test. To compare grades between groups before and after the intervention, we used the Mann-Whitney test. To test whether the grades before and after the intervention differed between groups, we used the Wilcoxon test. The software used for the analysis was R 3.0.1.\n\n\nResults\n\nThe objective of applying two tests about the topics of Cell Biology was to analyze the impact of self-experiments under conditions that mimic everyday situations – fasting, ingestion of water and carbohydrate oral solution ingestion (glucose or fructose) – on the achievement of university students of health sciences. In the test that preceded the experiment carried out in the present study, the grades of the students that participated actively and of those that only observed were similar (p > 0.05). In the test applied after the experiment, both groups reached higher grades (p < 0.01), but the individuals that participated actively in the experiment obtained higher grades than those that only observed it (p < 0.05; Figure 1 and Table 1). Test results for each student are available as Underlying data21.\n\nComparison of achievement in tests between students who participated actively in the experiments and students who only observed them. p < 0.05.\n\n1 Mann-Whitney test. SE, standard error.\n\nConsidering the differences between the subgroups of the students who participated in the experiment, there was no significant difference in grades among the subgroups fasting, glucose, and fructose (p > 0.05). Considering the differences within subgroups, only the students of the glucose and fructose groups showed an increase in their grades after the experiment (p < 0.05; Figure 2, Table 2).\n\nComparison of achievement in tests by students who participated actively in the experiment among the groups fasting, glucose, and fructose : p < 0.05.\n\n†Unless indicated.\n\nQuantitative variables (median, interquartile range).\n\nOver 90% of the students answered that learning is deeper when they have practical classes, that they are more important than theoretical classes, and that they are essential for learning. Over 70% of the students stated having high motivation. There was no difference in the distribution of answers in the questionnaire between the individuals that participated actively in the experiment and those that only observed it (Table 3). Raw survey results are available are available as Underlying data21.\n\n*By Fisher’s exact test.\n\n\nDiscussion\n\nThis study showed that students that provide blood samples for biochemical experiments in practical classes obtained higher grades than students that only observed the experiment. After the application of this teaching method, the group of students that participated actively in the practical classes (by providing blood samples) obtained higher grades than the group of students that only observed the experiment. However, both groups increased their grades, as shown by the learning assessment tests. This increase in learning corroborates the results generated by several authors, which indicate the efficacy of active-learning methods and several studies that showed an increase in learning through the use of active learning22–25. Nevertheless, most studies whose authors affirm that active-learning methods are better than others are based on theory, as they do not show experimental results or do not use proper control groups26. Several studies with a proper control group did not observe differences in learning between active-learning methods and passive, traditional methods4,27,28. Finally, some studies show that the use of active learning methods can even result in lower grades29,30.\n\nThese differences in results obtained with different learning methods suggest that other factors affect the efficacy of these methods and can occur regardless of the teaching method being more or less active. One of those factors seems to be the teacher’s behavior. For example, a teaching improvement tends to occur when teachers propose problems and questions, and stimulate a working atmosphere that give students freedom to take risks, make mistakes, and learn from their own mistakes31. Indeed, the attractiveness of educational activities developed is higher in the physical presence of the teacher32,33. Hence, we expect the learning strategies that provide the best learning atmosphere to be also the most efficient. The most striking result of the present study was the higher increase in grades of the group of students that participated actively in practical classes compared to the group of students that only observed them. This finding corroborates positive results of methods that involve self-assessment of learning by the students15. To our knowledge, this is the first study that shows that a teaching method in which the students assess their own body promotes deeper learning. Other studies have already showed that classes involving education of students using their own body as a subject can bring benefits19,18, but s no study has used control groups. One explanation for this phenomenon would be that this method ensures deep compulsory involvement of the student, which could increase their degree of attention and focus in the experiment and avoid distractions.\n\nThe results obtained by the students during the practical class could show them how the human organism responds to common everyday situations. In the groups that ingested carbohydrates, there was an increase in blood sugar and this increase was higher in the group that ingested glucose than in the group that ingested fructose. Similar to our experiment, several studies intended to measure blood sugar at the ingestion of glucose and fructose solutions34,35,36, and the everyday diet of people19,18. The lack of difference in grades in the test after the experiment among the subgroups that participated actively in the experiment (fasting, glucose, and fructose) indicates that the collection of one’s blood increases information comprehension and retention about the entire practical class and not only the result of one’s own organism or one’s own group. Hence, those that ingested glucose understood what occurs when one ingests fructose or remains fasting. This method may be more attractive to the students and may drive their attention The data of the analysis of the three groups that participated in the experiment also corroborate that. The significant increase in the grades only in the subgroups that ingested glucose and fructose and not in the grades of the subgroup that remained fasting suggests that the ingestion of solutions helped the students reach deeper learning compared to the students who only donated blood. The method used in the present study involves different forms of active participation and at the same time a study about the students’ own bodies at everyday situations and provides an active visceral learning.\n\nSurprisingly, the motivation and satisfaction in doing experiments, which was measured through a questionnaire, did not differ among subgroups. This result was unexpected, as more active learning methods tend to generate more interest1,2. However, the lack of difference may be due to the participation, as observers, of the students that did not use their own bodies in the practical classes. This explanation is corroborated by the fact that over 90% of the students answered that practical classes are more interesting and provide deeper learning. Other studies showed different values. For example, although medical students and residents subjected to active learning are more engaged and interact more with each other (43%) and with the teacher (57%), they reported lower perception of the objectives of learning4. In another study, most medical students studying clinical biochemistry (73%) opined that an active-learning method was more motivating, aroused the interest in the subject, and increased learning6. Another explanation for the results of our study is a possible insensitivity of the questions used in the questionnaire to subtle perceptions of students. Maybe the application of qualitative questions could better define the relationship between motivation and teaching method.\n\nWe conclude that active learning, in which the students use their own organism and blood in practical classes, can increase their grades in knowledge tests. This method led to an increase in learning compared to the control group, which only observed the experiments. In addition, the analysis of the three subgroups that participated actively in the experiment showed that the students who ingested glucose and fructose solutions had a significant increase in their grades, whereas those students who remained fasting did not. Such a difference suggests that a method that combines two different strategies, active learning and self-assessment, increases attention. The teaching tool showed in the present study is a positive alternative for university students in health sciences.\n\n\nData availability\n\nFigshare: Eduardo et al datasheet. https://doi.org/10.6084/m9.figshare.9725522.v221.\n\nThis project contains the following underlying data:\n\nEduardo et al.2019 dataset2 observers.xls.xlsx (survey and test results for each student).\n\nEduardo et al.2019 dataset1.xls (blood glucose levels before and after the experiment for each student actively participating in self-experimentation)\n\nFigshare: Eduardo et al datasheet. https://doi.org/10.6084/m9.figshare.9725522.v221.\n\nThis project contains the following extended data:\n\nQUESTIONNAIRE.docx (English translation of the questionnaire given to each student).\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).", "appendix": "Acknowledgments\n\nThe authors thank all participants in the study for assistance with data collection and Faculdade Anhanguera de Brasilia (FAB) and CAPES for support during the study.\n\n\nReferences\n\nEverly MC: Are students’ impressions of improved learning through active learning methods reflected by improved test scores? Nurse Educ Today. 2013; 33(2): 148–51. PubMed Abstract \| Publisher Full Text\n\nGhosh S, Dawka V: Combination of didactic lecture with problem-based learning sessions in physiology teaching in a developing medical college in Nepal. Adv Physiol Educ. 2000; 24(1): 8–12. PubMed Abstract \| Publisher Full Text\n\nGoldhaber DD, Brewer DJ: Does teacher certification matter? High school teacher certification status and student achievement. Educational Evaluation and Policy Analysis. 2000; 22(2): 129–145. Publisher Full Text\n\nHaidet P, Morgan RO, O'Malley K, et al.: A Controlled Trial of Active Versus Passive Learning Strategies in a Large Group Setting. Adv Health Sci Educ Theory Pract. 2004; 9(1): 15–27. PubMed Abstract \| Publisher Full Text\n\nEdelbring S, Broström O, Henriksson P, et al.: Integrating virtual patients into courses: follow-up seminars and perceived benefit. Med Educ. 2016; 46(4): 417–25. PubMed Abstract \| Publisher Full Text\n\nJoshi KB, Nilawar AN, Thorat AP: Effect of case based learning in understanding clinical biochemistry. International Journal of Biomedical and Advance Research. 2014; 5(10): 516–518. Publisher Full Text\n\nKulak Ve, Newton G: Um guia para usar baseado em casos de aprendizagem no ensino de bioquímica. Biochem Mol Biol Educ. 2014; 42: 457–473. PubMed Abstract \| Publisher Full Text\n\nParikh A, McReelis K, Hodges B: Student feedback in problem based learning: a survey of 103 final year students across five Ontario medical schools. Med Educ. 2001; 35(7): 632–636. PubMed Abstract \| Publisher Full Text\n\nPrince KJ, Van Eijs PW, Boshuizen HP, et al.: General competencies of problem-based learning (PBL) and non-PBL graduates. Med Educ. 2005; 39(4): 394–401. PubMed Abstract \| Publisher Full Text\n\nStevenson FT, Bowe CM, Gandour-Edwards R, et al.: Paired basic science and clinical problem-based learning faculty teaching side by side: do students evaluate them differently? Med Educ. 2005; 39: 194–201. PubMed Abstract \| Publisher Full Text\n\nDobson JL, Linderholm T: The effect of selected “desirable difficulties” on the ability to recall anatomy information. Anat Sci Educ. 2014. PubMed Abstract \| Publisher Full Text\n\nLarsen DP, Butler AC, Roediger HL 3rd: Repeated testing improves long-term retention relative to repeated study: a randomised controlled trial. Med Educ. 2009; 43(12): 1174–81. PubMed Abstract \| Publisher Full Text\n\nRivers WP: Autonomy at all costs: An ethnography of metacognitive self-assessment and self-management among experienced language learners. The Modern Language Journal. 2001; 85(2): 279–290. Publisher Full Text\n\nYang WM, Xu XZ: Self assessment in second-language learning. USChina Foreign Language. 2008; 6(56): 20–24.\n\nAgrawal S, Norman GR, Kevin WE: Influences on medical students’ self-regulated learning after test completion. Med Educ. 2012; 46(3): 326–35. PubMed Abstract \| Publisher Full Text\n\nHarasym PH, Tsai TC, Munshi FM: Is problem-based learning an ideal format for developing ethical decision skills? Kaohsiung J Med Sci. 2013; 29(10): 523–9. PubMed Abstract \| Publisher Full Text\n\nKhan BA, Ali F, Vazir N, et al.: Students' perceptions of clinical teaching and learning strategies: a Pakistani perspective. Nurse Educ Today. 32(1): 85–90. PubMed Abstract \| Publisher Full Text\n\nPassos RM, Sé AB, Wolff VL, et al.: Pizza and pasta help students learn metabolism. Adv Physiol Educ. 2006; 30(2): 89–93. PubMed Abstract \| Publisher Full Text\n\nHopper MK, Maurer LW: Laboratory exercise: study of digestive and regulatory processes through the exploration of fasted and postprandial blood glucose. Adv Physiol Educ. 2013; 37(3): 254–263. PubMed Abstract \| Publisher Full Text\n\nConsorti F, Mancuso R, Piccolo A, et al.: Evaluation of the acceptability of Peer Physical Examination (PPE) in medical and osteopathic students: a cross sectional survey. BMC Medical Education. 2013; 13: 111. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEduardo AM, Rosa EC, Welker AF: Eduardo et al datasheet. figshare. Dataset. 2019.\n\nCruz SC, Carvalho AA: Produção de Vídeo com o Movie Maker: Um Estudo Sobre o Envolvimento dos Alunos de 9.º. Ano na Aprendizagem. 2007. Reference Source\n\nErwin TD, Rieppi R: Comparing multimedia and traditional approaches in undergraduate psychology classes. Teaching of Psychology. 1999; 26: 58–61. Publisher Full Text\n\nNarula N, Ahmed L, Rudkowski J: An evaluation of the '5 Minute Medicine' video podcast series compared to conventional medical resources for the internal medicine clerkship. Med Teach. 2012; 34(11): e751–5. PubMed Abstract \| Publisher Full Text\n\nWiecha JM, Chetty VK, Pollard T, et al.: Web-based versus face-to-face learning of diabetes management: the results of a comparative trial of educational methods. Fam Med. 2006; 38(9): 647–52. PubMed Abstract\n\nNascimento GP: Estudo Controlado da Efetividade de um Instrumento que Acopla Aprendizagem Ativa e Tecnologia: Criação de Vídeos pelos Estudantes. Dissertação de Mestrado. UnB. Brasília. 2014. Reference Source\n\nChao SH, Brett B, Wiecha JM, et al.: Use of an online curriculum to teach delirium to fourth-year medical students: a comparison with lecture format. J Am Geriatr Soc. 2012; 60(7): 1328–32. PubMed Abstract \| Publisher Full Text\n\nSchreiber BE, Fukuta J, Gordon F: Live lecture versus video podcast in undergraduate medical education: A randomised controlled trial. BMC Med Educ. 2010; 10: 68. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrewer DJ, Goldhaber D: Why don’t schools and teachers seem to matter? J Hum Resour. 1997; 32(3): 505–523. Publisher Full Text\n\nSchwerdt G, Wuppermann AC: Is traditional teaching really all that bad? A within-student between-subject approach. Econ Educ Rev. 2011; 30(2): 365–379. Publisher Full Text\n\nBiggs J, Tang C: Teaching for quality learning. at university: What the student does. McGraw-Hill Education (UK), 2011. Reference Source\n\nAli WG: Caring and Effective Teaching Behavior of Clinical Nursing Instructors in Clinical Area as Perceived by Their Students. J Educ Prat. 2012; 3(7): 15–27, 2012. Reference Source\n\nWarwick P, Mercer N, Kershner R, et al.: In the mind and in the technology: the vicarious presence of the teacher in pupil’s learning of science in collaborative group at the interactive whiteboard. Comput Educ. 2010; 55: 350–62. Publisher Full Text\n\nAbraha A, Humphreys SM, Clark ML, et al.: Efeito agudo da frutose na lipemia pós-prandial em indivíduos diabéticos e não-diabéticos. British Journal of Nutrition. 1998; 80: 169–175. PubMed Abstract \| Publisher Full Text\n\nBidwell AJ: Impact of Physical Activity on Postprandial Lipidemia and Glycemic Control after Chronic Fructose Ingestion. Exercise Science - Dissertations. 2012; 8. Reference Source\n\nZawiasa A, Nowicki M: Acute effects of fructose consumption on uric acid and plasma lipids in patients with impaired renal function. Metabolism. 2013; 62(10): 1462–1469. PubMed Abstract \| Publisher Full Text" }	[ { "id": "62046", "date": "15 Apr 2020", "name": "Dakota Armour", "expertise": [ "Reviewer Expertise Medical education research" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAn important strength of this article is that it shows how teachers can help students learn a potentially dull topic actively by making scientific concepts personally relevant and interesting. There is growing interest in the application of embodiment theory in the field of health professions education so using students’ own bodies as a learning tool, which is a topically relevant feature of the intervention, is another strength of the article.\n\nThere are, though, threats to the validity of the research, which prevent us agreeing with the authors’ claim that their ‘teaching method in which students assess their own body promotes deeper understanding’. The authors took a classical scientific experimental approach but there were shortcomings in various aspects of the research design:\nThe authors did not frame a null hypothesis. This is a shortcoming because classical experiments depend on enabling researchers to reject a null hypothesis: in reality, we could not find any clearly-articulated hypothesis, which is particularly important when taken together with comment 4, below.\n\nBlinding as to whether a participant is subjected to the experimental or the control condition is an important feature of experiments. In reality, participants in this research knew what group they were in and observed members of other groups. It is plausible that this knowledge influenced participants’ mindsets and confidence levels and plausible, therefore, that this influenced the outcomes of the post-intervention tests. We have no way of knowing how strong that influence may have been, and even in which direction, but it may well have confounded the results of the experiment.\n\nThe randomisation protocol is not described and the information available makes us question the rigour with which this was done.\n\nThe authors do not appear to have defined, before the experiment, which statistical comparisons would allow them to draw inferences. As a result, they appear to have relied on post-hoc comparisons, and used the 5% level as an absolute arbiter of difference or lack of difference. The snag is that these comparisons were confounded by small but potentially important differences in baseline test scores, which directionally favoured the intervention over the control condition. The authors are at risk of Type II statistical error (false negative) when they conclude that there was no difference between the active and control groups in the pre-intervention test scores because, with larger numbers of subjects, there could have been. In reality, the difference in scores between groups was 0.1>p>0.05, which is not much above the 5% level that is deemed significant.\n\nWe suggest that a comparison between absolute difference in scores from pre to post intervention, adjusted for the confounding effect of baseline levels, would have been a more appropriate analysis and that, at least with the numbers included in this experiment, no statistically significant difference would have been found.\n\nFigure 1 appears to contradict findings stated in the text and we found it hard to understand.\n\nWe would expect any empirical article to include, in the discussion, consideration of the strengths and limitations of the study.\n\nWe could not work out if there were two questionnaires - one assessing knowledge - or just a questionnaire assessing motivation and satisfaction? Wording needs clarification- ‘The assessment of students’ knowledge levels occurred through two written, closed-notes and closed-book tests and the application of questionnaires on the topics in the classroom’. If questionnaires and tests assessed knowledge, detail is needed to illustrate how data from the two sources were amalgamated.\n\nNo detail was given about the questionnaire on satisfaction and motivation.\n\nThere are two typos on page 7. Paragraph one ‘but s no’ and paragraph two missing full stop ‘may drive their attention’.\n\nConclusion: On methodological grounds, we feel the authors cannot reject the null hypothesis that their intervention was ineffective. That is a shame because it might have been! It is more plausible that an active learning approach would be of educational value than that it would not be of educational value, but the study did not show this.\n\nWe suggest that the authors were not just let down by flaws in their research design, but by their decision to subject education to classical scientific methodology. Regehr1 has urged education researchers to move from the ‘imperative of proof’ (which was unsuccessfully applied here) to the ‘imperative of representing complexity well’, which the authors’ intervention, but not their evaluation, sought to do. Education is not so simple as comparing a pill and a placebo. Learners are complex, their subject matter is complex, and learning environments are complex.\n\nThe education community is waking up to the fact that classical scientific experimentation, which is predicated on reducing complex systems to intervention-control comparisons, is ill-suited to complex social systems like education. Qualitative research, in the right hands, can examine relationships between conditions, mechanisms and outcomes in complex learning environments.2 So whilst we are still in the dark about whether this particular intervention was effective, the study highlighted the need for education researchers both to conduct their work rigorously, and to extend their methodological skills-set.\n\nDakota Armour (medical student) and Tim Dornan (doctor and education researcher)\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "62044", "date": "16 Apr 2020", "name": "Fabrizio Consorti", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript deals with a very interesting and intriguing topic, as a special instance of the broader topic of active learning. Therefore, even if it is a small-scale trial, it deserves interest.\nThe main problem is the lack of theoretical foundation, that in my opinion causes an inconsistent interpretation of the results. Actually the authors saw a very good track of interpretation when they wrote “The method used in the present study involves different forms of active participation and at the same time a study about the students’ own bodies at everyday situations and provides an active visceral learning.” (page 7, first column). I suggest that they should dig into this topic, using the concepts and theoretical background of the “embodiment” to drive the Discussion.\nI suggest some reading at this proposal, just as examples (I’m not the author of any of these articles…) to focus the idea of “learning with/about own body”:\nBody pedagogics: embodied learning for the health professions. Kelly M, Ellaway R, Scherpbier A, King N, Dornan T. Med Educ. 2019 Oct;53(10):967-977. doi: 10.1111/medu.139161\n\nOn the Education About/of Radical Embodied Cognition. van der Kamp J, Withagen R, Orth D. Front Psychol. 2019 Nov 5;10:2378. doi: 10.3389/fpsyg.2019.02378. eCollection 2019.2\n\nNeuroscience of Virtual Reality: From Virtual Exposure to Embodied Medicine. Riva G, Wiederhold BK, Mantovani F. Cyberpsychol Behav Soc Netw. 2019 Jan;22(1):82-96. doi: 10.1089/cyber.2017.29099.gri. Epub 2018 Sep 5.3\n\nLearning through the senses. van der Schaaf M. Med Educ. 2019 Oct;53(10):960-962. doi: 10.1111/medu.13943.4\n\nThe lack of theoretical foundation is also reflected by some of the statement of the manuscript:\nPage 3 , 1st column - “Problem-based learning (PBL), which frequently involves simulations of real situations”: that’s Case-based Learning (CBL), another active method.\n\nPage 5, first paragraph of the Discussion: the authors quote single studies to support they claim about the inconclusiveness of research about the superiority of active learning methodologies, but we have plenty of high-grade evidence about this topic. When meta-analyses are available, they should be quoted in the Discussion, instead of a “cherry-picking” approach to low grade evidence. I list just some, as an example:\nEffectiveness of various innovative learning methods in health science classrooms: a meta-analysis. Kalaian SA, Kasim RM.Adv Health Sci Educ Theory Pract. 2017 Dec;22(5):1151-1167. doi: 10.1007/s10459-017-9753-6.5\nThe effectiveness of the problem-based learning teaching model for use in introductory Chinese undergraduate medical courses: a systematic review and meta-analysis. Zhang Y, Zhou L, Liu X, Liu L, Wu Y, Zhao Z, Yi D, Yi D.PLoS One. 2015 Mar 30;10(3):e0120884. doi: 10.1371/journal.pone.0120884.6\nDoes problem-based learning work? A meta-analysis of evaluative research. Vernon DT, Blake RL.Acad Med. 1993 Jul;68(7):550-63. doi: 10.1097/00001888-199307000-00015.7\nThistlethwaite JE, Davies D, Ekeocha S, et al. The effectiveness of case-based learning in health professional education. A BEME systematic review: BEME Guide No. 23. Med Teach. 2012;34(6):e421–e444. doi:10.3109/0142159X.2012.680939.8\n\nPage 5, second paragraph of the Discussion: I cannot understand why here the authors discuss the role of teacher, an element that had no relevance in their experiment. Further, the references supporting this paragraph are of studies on high school students.\nSome more comments:\nAbstract (and also in the text, at the end of the Introduction): the goal of the study is described as “to assess whether this intervention could help assess the influence of self-experiments on student learning.”. I suggest that the goal of the study should be better expressed with reference to a research hypothesis “we hypothesize that the intervention will produce a better … than …”. So, the goal is to confirm or refuse the hypothesis\n\nPage 3, first column “The few reports on self-experiments frequently involve activities that are not part of the students’ routine”. There is only one reference, so either you add “as for example in…” or you add some more references. The same problem is at page 5, end of 1st column “Nevertheless, most studies whose authors affirm that active-learning methods are better than others are based on theory, as they do not show experimental results or do not use proper control groups (26).” I also highlighted “most studies”, because it is a statement that should be supported by data. Only a systematic review could support this statement. See also my comment about the use of high-grade evidence in Discussion.\n\nExperimental design, page 3, 2nd column: second line: “dived” for divided “According to a sample calculation performed using G*Power 3.1 software, the group should contain a total of 71 students of the second term the Bachelor of Pharmaceutical Sciences (33 active and 38 observers 38) to achieve a statistical power of 60% to achieve the objectives.” I cannot understand because the groups are different in number of students.\n\nFigure 1: I guess that the labels of the columns are inverted, because the score is higher before than after. The vertical axis is not clear, it seems that the students gave only 8% of right answers.\n\nTable 1, second cell: “self-experimentation”, please correct.\n\nTable 2: I find no use in intertwining the data of glycemia with the descriptive data. They are not experimental data. On the contrary, the descriptive data of the control group are missing, so the reader can only guess that the two groups were comparable for sex and age.\n\nDiscussion, page 7, 1st column, “The significant increase in the grades only in the subgroups that ingested glucose and fructose and not in the grades of the subgroup that remained fasting”. This statement seems to be contradictory with what stated few lines before “The lack of difference in grades in the test after the experiment among the subgroups that participated actively in the experiment (fasting, glucose, and fructose)”.\n\nDiscussion, page 7, beginning – “This finding corroborates positive results of methods that involve self-assessment of learning by the students (15).” In the international literature of medical education (and in reference 15) self-assessment is intended as evaluation of outcomes of learning by the students themselves and not as measuring something upon themselves, like the glycemia.\n\nFew lines below “But s no study has used control groups”, please cancel the “s”.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-1623
https://f1000research.com/articles/8-1622/v1	11 Sep 19	{ "type": "Research Article", "title": "Understanding life sciences data curation practices via user research", "authors": [ "Aravind Venkatesan", "Nikiforos Karamanis", "Michele Ide-Smith", "Jonathan Hickford", "Johanna McEntyre", "Nikiforos Karamanis", "Michele Ide-Smith", "Jonathan Hickford", "Johanna McEntyre" ], "abstract": "Background: Manual curation is a cornerstone of public biological data resources. However, it is a time-consuming process that urgently needs supportive technical solutions in the face of rapid data growth. Supporting scalable curation is a part of the mission of the Elixir Data Platform. Thus far, we have established infrastructure capable of ingesting and aggregating text-mined outputs from multiple providers and making these available via an API. This public API is used by Europe PMC to display specific entities and relationships on full text articles (via the SciLite application). Methods: To ensure that the future development of this infrastructure meets the needs of curators, we carried out a user research project to understand and identify common workflow patterns and practices via an observational study. Building on these outcomes, we then devised a curator community survey to more specifically understand which entity types, sections of a paper and tools are of top priority to address. Results: The main challenges faced by curators included the following: a) There is a need for ways to prioritise and identify relevant papers for curation as the volume of literature is large; b) Finding specific information can prove difficult; quick ways of filtering articles based on specific entities, such as experimental methods, species and other important entities, such as genes, cell lines and tissue samples, are required; and c) Transferring information from the search/annotation tools to the various curation workflows was also challenging. Conclusions: This study lays the foundation for identifying actionable items to orient the current infrastructure towards meeting the needs of curation community, by improving text-mined annotation quality and coverage and other engineering solutions; and reusing text-mined annotations and other metadata in Europe PMC for article triage. Furthermore, this study presents an opportunity to explore customisation of triage/ranking systems to suit different curation contexts.", "keywords": [ "Database curation", "User research", "Observational study", "Curator survey", "Annotation Infrastructure", "Europe PMC" ], "content": "Introduction\n\nBiological databases play a key role in knowledge discovery in life science research. A major contributor towards the maintenance of these databases is the process of manual curation. Curation is a high value task as experts carefully examine the relevant scientific literature and extract the essential information, such as biological functions and relationships between biological entities, generating the corresponding database records in a structured way. The advances in high-throughput technologies have resulted in tremendous growth of biological data, consequently increasing the number of research papers being published. As a result, the demand for high-quality curation that makes use of these resources has never been higher, but this demand can present challenges for curators in finding, and assimilating scientific conclusions described in the literature.\n\nText mining, machine learning and analytics promise to provide better ranking of reading lists, classification of articles, and identification of assertions with their biological context and evidence buried within the text of articles. To this end, many life science knowledgebases now include text mining (to varying degrees) in curation workflows. For example, databases, such as neXtProt1,2 and FlyBase3 have integrated text mining algorithms into their respective curation workflows to retrieve a ranked list of relevant articles and tag entities of interest. Furthermore, tools like PubTator4 and TextPresso5 are other examples of text mining tools that have been adopted by some curation communities. On the other hand, databases that mainly rely on manual curation, such as IntAct6 and DisProt7, are exploring possibilities to leverage text mining approaches to select articles for further curation.\n\nBroadly speaking, the curation community recognises the potential of text mining in article triage and the identification of entities/concepts for curation. Nevertheless, text mining pipelines adopted thus far have been engineered to cater to specific domains or projects and wide uptake is lacking; curators often continue to use manual curation methods. This mainly stems from the wide variety of very precise information required by curators. The challenge is therefore to produce robust systems that both address the immediate and specific needs of curators as well as scale across multiple curation groups. In order to do this, we need to know the immediate challenges faced by curators with respect to selection and prioritisation of articles to curate. A clear understanding of the requirements will help build new systems and/or re-orient existing systems that cater to the needs of the curation community.\n\nIn this report we describe the outcomes of a user research project, conducted to understand curation practices and priorities for article selection. The project comprised of two parts, a) an observational study, to understand how curators proceed with selecting articles to curate, to identify commonalities in curator requirements; and b) a community survey, to specifically identify the immediate priorities of curators, such as entity types and sections of interest in an article, to name a few. The aim of this study is to identify specific actions for the Elixir Data Platform in the future, optimising and extending existing systems and infrastructural components. In the subsequent sections we present the main findings from our investigation.\n\n\nMethods\n\nWe initially drafted an interview guide (list of questions available as Extended data8) and a preliminary curator persona, reflecting our initial hypotheses about curators and their work practices. Following this we selected five different curation teams. The selection criteria was based on the type of curation the group was involved with, i.e., extracting biological information from scientific literature and integrating it into a biological database, contrary to groups that process raw data submissions (such as sequencing data). Out of these, two teams were based at EMBL-EBI, the other three teams were based in Norway, Switzerland and Italy. For teams that were situated at EMBL-EBI, the interviews were held in person and for the other teams the interviews were conducted over conference calls. The sessions were conducted over a period of two months: between March and April 2018. We observed three project leaders and four curators from the selected teams. The participants belong to teams that focused on curating very specific experimental evidence primarily on proteins such as protein-protein interactions, the role of a protein in a complex, protein disruption, human protein functions and transcription factor regulation. One team is focussed on the annotation of human genes relevant to a particular disease and another one on curating publications reporting associations of genetic variants with diseases.\n\nFor each session we followed an iterative user research process9,10 (as outlined in Table 1):\n\nThe table provides an overview of the observation study on curation practices.\n\nThe participants were asked to proceed with their daily curation work11 and were observed on how:\n\n○ they select the entities (either come from a spreadsheet or partially curated data record) they wish to curate,\n\n○ perform searches (including the query parameters) to retrieve the initial set of publications,\n\n○ the criteria they used to either discard or select an article,\n\n○ the information from the selected article is transferred to the respective curation platform.\n\nUsing the “What? So What? Now What?” method1, we transcribed our notes from these sessions and identified the most important observations, patterns and their implications.\n\nAdditionally, we further carried out two follow-up interviews with one project lead and one curator from the same team at EMBL-EBI to clarify particular curation tasks.\n\nFurthermore, we conducted a stakeholder workshop12 with three project leads and four curators from four different curation teams to validate our main learnings from the interview sessions. As some curators took part both in the interviews and the workshop, overall we have engaged with 12 curators and team leads from seven different teams. The participants were presented with the preliminary curator persona13 and a workflow outlining the curation process. The participants were invited to give us their feedback on these drafts and express their challenges or pain points as How Can We (HCW) questions2. Their feedback was used to revise the curator persona and the curation process workflow and was consolidated into the curation experience map.\n\nBased on the interview guide used for the observational study, we formulated questions to understand the immediate challenges. The survey consisted of 15 questions (see Extended data8) ranging from, for instance, the section of the article the curators were most interested in; the types of biological entities curators look for; and whether it helps to know that a given article has been curated/accounted for in another database. The community survey was conducted online and was developed using Typeform. The survey was promoted via the mailing lists of various consortia, such as ELIXIR, International Society for Biocuration (ISB) and Alliance of Genome Resources. These widely known consortia provide a forum for developers, researchers and curators to streamline and standardise the maintenance of biological resources. The survey was conducted between December 2018 to January 2019.\n\nWe confirm that we have obtained consent to use data from the participants as per the Europe PMC privacy notice. The privacy notice is formulated in accordance with EMBL’s data protection framework. The consent was part of the survey form and the participants can take the survey on accepting the terms and conditions for data re-use.\n\n\nResults and discussion\n\nCurator persona. We created a persona called Ashley (see Underlying data8) to present the curators’ needs, sentiments, tasks and pain points from their own perspective in more detail to help us empathise with them13:\n\nAshley curates with precision and attention to detail, while trying to be as efficient as possible. Ashley is looking for very specific information about an experiment that the authors of a paper do not always report in a lot of detail. Ashley appreciates being able to ask a team mate when the “detective work” bears no fruit.\n\nApart from these “organic, informal discussions”, Ashley works independently during “triage”, “annotation” and to fill in the curation record in the Editor. During the latter stage Ashley tries to “translate from the author’s language to the curator’s language” using the appropriate identifiers and Control Vocabulary (CV) terms for species, proteins, methods and other important entities so that the curated evidence is referred to precisely and consistently and the annotations in the curation record are self explanatory outside the context of the paper.\n\nThis is cumbersome as a particular type of evidence is not always referred to in the same way and in enough detail in the literature. Moreover, the Editor is not integrated with the search and annotation tools, so Ashley spends a lot of time going back and forth between the paper and the Editor, translating the curatable text from the paper into CV terms, switching browser tabs and consulting notes from the online research and team discussions.\n\nCuration experience map. The curation experience map in Figure 1 presents the identified pain points in the context of the main curation activities. As shown in the map, curation consists of four stages:\n\na) Deciding which entity (primarily protein in this case) to curate. What to curate often depends on the curator’s background and the project that they are working on.\n\nb) “Triaging” the literature to identify relevant publications.\n\nc) “Annotating” a relevant publication to identify the precise curatable information in detail, including determining the species and the relevant experimental method.\n\nd) Filling in the curation record based on the curatable information in the publication (which is often done in parallel with annotating the paper).\n\nIn a typical curation scenario, curators:\n\nSearch PubMed for a protein and scan the titles in the search results to identify relevant experimental papers during “triage”.\n\nIf a title indicates that the paper is relevant (e.g. by mentioning the protein and/or species of interest), then they skim read the Methods and the Results of the paper. They are particularly interested in Figures, Tables and their Legends, which is where they usually find the key (curatable) information.\n\nThese sections are read more thoroughly during the “annotation” stage to identify the exact experimental context that needs to be curated. They may “glance through” the Abstract or skip it altogether.\n\nA pain point during “triage” is identifying relevant publications: The curators reported that most publications returned in a PubMed search are usually not relevant (i.e. they are false positives). Additionally, because they are often looking for very specific and at times underreported experimental evidence, some searches return very few papers or no papers at all.\n\nAmbiguities in the paper about species, proteins, and relevant experimental methods may slow down curation significantly during “triage” and “annotation”. The curators highlighted that identifying the species as their main pain point as this task may take up to “75% of the curation effort”, and in the end may turn out to be irretrievable from the paper.\n\nFurthermore, to get clarity on the details of an experiment, curators would look up specific references in the paper and do further research online. If this “detective work” is not successful, the curators will either not annotate the paper or will provide fewer annotations. The curators would also discuss unresolved questions such as the annotation of unusual data or what to do when there is no curatable data with their teammates during “organic, informal discussions”. As a last resort they would contact the authors directly; however, authors often do not respond to requests for clarification.\n\nIf there are no matching CV terms to annotate the paper new ones are requested. This can delay annotating the paper because the ontology staff are often different to those curating the papers and sometimes requesting new terms results in prolonged discussions between curators and ontologists.\n\nIt was observed that curators use different tools at each stage, which are not integrated with each other. During “triage” they would search PubMed for relevant publications and then look at a particular paper on the publisher’s site. Annotating a publication may involve downloading or printing the pdf version of the paper and highlighting curatable text. To fill in the curation record they use a bespoke tool which they call “the Editor”, which presents a template that needs to be filled in with molecule names and experimental context, supported by standardised identifiers, controlled vocabularies and ontologies as well as free text describing the experimental evidence. Most of the times the assertions that go in the database are not in the paper in the same words.\n\nSummary of the pain points in curation workflows. The identified pain points were formulated as How Can We (HCW) questions as follows:\n\nHCW identify relevant publications for curation in search results or a list of references during “triage”?\n\nHCW identify species, experimental methods, molecules (primarily proteins) and other important entities (such as cells and tissues) in a publication during “triage” and “annotation”?\n\nHCW help curators fill in the curation record more efficiently?\n\nThe survey received 42 respondents in total, covering a number of European countries, such as the United Kingdom, France, Italy and Switzerland. The majority of the participants identified themselves as ‘Scientific curator’ with over 5 years experience in curation. Broadly speaking respondents mainly curate peer-reviewed articles (43.6%), followed by review (25.5%) and preprint (16%) articles. Figure 2 shows their preference in the type of articles for curation.\n\nFigure 3 shows the section of articles of interest to the curators: the majority look for method sections in articles. The other sections of importance are the figures/tables and their legends. Apart from these, the supplementary data seems to be a section of importance. Furthermore, as shown in Figure 4, the types of entities curators look for in articles were diverse with preference given mainly to: genes/protein curation and their functions, database accession numbers, experimental methods and gene mutations.\n\nRespondents were asked if it was useful to know a given article was already curated by another database: the results indicate that it was useful (see Figure 5). A follow-up question was asked as to (if useful) why it was useful to know if an article is already annotated by another database, the majority of the responses ranged from: avoiding duplication if the curators belong to the same consortium, as a means of validation (in case of ontology terms), and consistency in annotations.\n\nText mining approaches are sophisticated and play a vital role in addressing big data questions, the results of which can contribute to supplying “leads” on key papers for curation. However, curators require a wide variety of very precise information. Addressing each of those specific requirements will be a complex task, but text mining systems can certainly provide underlying services based on broad commonalities in the requirements that can prove useful to curators. To this end, this effort has proved to be useful in terms of understanding the main challenges faced by curators. While the sample size of the community survey was small, when analysed in conjunction with the observational study we found significant commonalities on the work practices. For instance, in the survey when the respondents were asked for their biggest challenge while curating, the majority responses indicated finding relevant papers, and identifying specific information that includes genes and species.\n\nOur research on curation practices so far indicates a need to better support curators on the following areas:\n\nIdentify relevant papers for curation during “triage”. An efficient way towards article selection, where search results could be prioritised based on a set of parameters.\n\nIdentify species, relevant experimental methods, molecules (primarily proteins) and other important entities (such as cells and tissues) in a publication during “triage” and “annotation”.\n\nRetrieving certain sections of articles such as Methods, Figures or Results.\n\nIntegrate triage systems to the various curation workflows.\n\n\nConclusion\n\nContributions made by manual curation are vital to the maintenance of biological databases. To maximise the impact of this critically important process, the latest technological advancements need to be leveraged. Under the Elixir Data platform, we have established infrastructural elements to support scalable curation, which includes automated systems to ingest and aggregate from various sources, APIs to redistribute the annotations and an application called SciLite to display annotations on articles. However, a key challenge for scalable curation is to make use of such core components across different curation teams, whose requirements and workflows can be highly precise and vary widely. Consequently, this requires engagement with the curation community to derive actionable insights that may contribute towards service delivery. Therefore, this project lays the foundation needed to understand the commonalities shared among various curation workflows. Going forward, we will use the results of the project to feed into improvements to text-mined annotation quality and coverage, triage and browsing systems or other engineering solutions.\n\n\nData availability\n\nThe interview responses from the observation study have not been made public to protect the participants’ privacy. Please contact the corresponding author to apply for access to the data, providing details of the information required and the intended use of the data. Access to the data will be granted once permission from participants to share the data has been obtained.\n\nZenodo: Results of user research project to understand data curation practices. https://doi.org/10.5281/zenodo.32096588.\n\nThis project contains the following underlying data:\n\nCurator persona.docx (the curator persona generated during the first part of the study).\n\nCurator survey results.xlsx (raw data taken from the survey given to each participant).\n\nZenodo: Results of user research project to understand data curation practices. https://doi.org/10.5281/zenodo.32096588.\n\nThis project contains the following extended data:\n\nObservation study - interview guide.docx (interview guide outlines the type of questions to be asked)\n\nCurator survey questions.docx (questionnaire given to each participant in the community survey).\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).", "appendix": "Acknowledgements\n\nWe are grateful to our participants, to Francisco Talo for his involvement and to Ane Møller Gabrielsen for her comments.\n\n\nFootnotes\n\n1http://www.myddelton.co.uk/blog/what-so-what-now-what\n\n2http://www.designkit.org/methods/3\n\n\nReferences\n\nLane L, Argoud-Puy G, Britan A, et al.: neXtProt: a knowledge platform for human proteins. Nucleic Acids Res. 2012; 40(Database issue): D76–D83. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMottin L, Gobeill J, Pasche E, et al.: neXtA5: accelerating annotation of articles via automated approaches in neXtProt. Database (Oxford). 2016; 2016: pii: baw098. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKaramanis N, Seal R, Lewin I, et al.: Natural language processing in aid of FlyBase curators. BMC Bioinformatics. 2008; 9: 193. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWei CH, Kao HY, Lu Z: PubTator: a web-based text mining tool for assisting biocuration. Nucleic Acids Res. 2013; 41(Web Server issue): W518–22. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMüller HM, Kenny EE, Sternberg PW: Textpresso: An Ontology-Based Information Retrieval and Extraction System for Biological Literature. PLoS Biol. 2004; 2(11): e309. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOrchard S, Ammari M, Aranda B, et al.: The MIntAct project--IntAct as a common curation platform for 11 molecular interaction databases. Nucleic Acids Res. 2014; 42(Database issue): D358–63. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPiovesan D, Tabaro F, Mičetić I, et al.: DisProt 7.0: a major update of the database of disordered proteins. Nucleic Acids Res. 2017; 45(D1): D219–D227. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVenkatesan A, Karamanis N, Ide-Smith M, et al.: Results of user research project to understand data curation practices. [Data set]. Zenodo. 2019. http://www.doi.org/10.5281/zenodo.3209659\n\nGothelf J, Seiden J: Lean UX: Designing Great Products with Agile Teams. O’Reilly Media, Inc. 2016. Reference Source\n\nKaramanis N, Pignatelli M, Carvalho-Silva D, et al.: Designing an intuitive web application for drug discovery scientists. Drug Discov Today. 2018; 23(6): 1169–1174. PubMed Abstract \| Publisher Full Text\n\nBeyer H, Holtzblatt K: Contextual Design: Defining Customer-Centered Systems. Morgan Kaufmann. 1997. Reference Source\n\nGray D, Brown S, Macanufo J: Gamestorming: A Playbook for Innovators, Rulebreakers and Changemakers. O’Reilly Media, Inc. 2010. Reference Source\n\nCooper A: The Inmates Are Running the Asylum: Why High Tech Products Drive Us Crazy and How to Restore the Sanity. Sams-Pearson Education. 2004. Reference Source" }	[ { "id": "53750", "date": "07 Oct 2019", "name": "Lynette Hirschman", "expertise": [ "Reviewer Expertise Evaluation of text mining for biomedical applications", "particularly curation." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a well-designed and informative examination of data curation practices in support of the Elixir Data Platform, with a focus on exploring curator needs and pain points to identify where automated tools might help. The article presents results from a series of studies based on interviews with curators and responses from a questionnaire filled out by over 40 curators from 11 countries.\n\nThese results are consistent with an earlier article from 2012, Text Mining for the Biocuration Workflow1, which summarized the findings of a workshop held at the third International Biocuration Conference. Below is a short from the 2012 paper:\n* Curators wanted tools that were easy to use, easy to install and easy to maintain by the intended end user (ideally, a developer associated with the curation team, who will not necessarily be an expert in text mining or natural language processing). The tools do not have to be perfect, but they need to complement (not replace) the biocurator's function. A number of curation groups indicated that they would use the tools to do an initial batch processing, followed by biocurator validation, where the biocurator makes a yes/no decision and avoids having to type or look names up in a large database. Another important use was linking mentions of biological entities in text with the correct identifiers in biological databases, as well as linkage to the appropriate ontology terms. A number of curators felt that they would like text mining tools to aid in identifying and prioritizing papers for curation, to avoid wasting time on papers that did not have ‘relevant’ (e.g. curatable or novel) results. They also wanted tools to identify the sections of full-text papers containing curatable information. *\nIt would be informative to do a comparison of the findings from the 2012 paper with this paper, to see whether curator needs have changed – and to what extent automated tools have been able to address some of the curator needs.\n\nOne of the paper’s most interesting findings from the curator feedback is the need to identify species. This was flagged as a particular pain point, taking ‘up to 75% of the curation effort’! This has also been a consistent stumbling block for text mining systems, because identification of species is essential to link a mention of a gene or protein to the correct accession number in databases such as UniProt or EntrezGene. This turns out to be a hard problem for text mining systems (as well as for curators) for several reasons: 1) mentions of species are often given as background information and may well not be mentioned in the same sentence (or even section) as mentions of the gene or protein being studied; 2) authors may want to generalize findings to other species (especially to humans) even when the experiments have been done on other species; 3) information about the specific experimental constructs may be buried in the methods section – and may involve inserting a gene from one organism into the genome of a different organism.\n\nOne interesting omission is a discussion of the need for interactive curation tools. This has been a major theme of recent BioCreative evaluations (see, e.g., Overview of the interactive task in BioCreative V2). If an interactive system could show the curator a prioritized list of candidates, this could speed up several curation activities. Specifically, an interactive system could, e.g., show a ranked list of candidate papers for curation, with evidence highlighted; or show a paper (section) with a gene or protein mention highlighted together with a selectable list of candidate species, so the curator could quickly select the correct species in order to link to the correct accession number; or show highlighted evidence sentences supporting protein-protein interaction, for quick validation or rejection by the curator. Given a goal of providing tools to speed the curation workflow, interactive systems offer a promising approach by putting the human in the loop to augment text mining, where automated tools do not, on their own, provide sufficient accuracy.\n\nFinally, in the Conclusion section, the authors discuss the need to tailor capabilities for different curation tasks or workflows. Identifying commonalities is indeed key, as the authors note; but there is also an urgent need to develop methods to quickly tailor tools to new tasks or specific requirements in a curation workflow. This is an underexplored area, but may be key to more widespread adoption of text mining tools.\n\nSpecific comments:\nAdd references to some of the older background work in this area.\n\nThe Conclusions section of the Abstract mentions “actionable items” but doesn’t provide specifics. The list on p. 9, col 1 top is informative, and could be included in the abstract, e.g., prioritizing papers; filtering articles based on specific entity types; and retrieving specific sections of articles.\n\nA table summarizing the different interactions with curation teams and curators would be helpful, along with the specific types of curation being done by the teams. This information is presented in the text at different points, but it is hard to keep track without a summary, since there were several rounds and types of interactions.\n\nIn Figure 3, it would be useful to know the denominator, as well as the actual number of curators identifying specific sections.\n\nFigure 4 is very useful, but hard to read. In addition, it would be interesting to know what specific controlled vocabularies are in use for each of these types of information.\n\nFigure 5 – again, it would be useful to know the denominator (the total number of respondents for that question).\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "53748", "date": "14 Oct 2019", "name": "Cecilia N. Arighi", "expertise": [ "Reviewer Expertise My area of expertise is on biocuration", "usability and text mining applied to biocuration." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis work presents a study about the biocuration community and its literature-based curation practices. The work intends to identify pain points and commonalities in the curation workflow where ePMC infrastructure could work to assist this community.\nThe study includes the detailed observation of 5 curation groups following their regular literature curation work and a survey targeted to the biocuration community to learn about the literature-based curation tasks. The approach is appropriate and well designed. However, it seems that the curation groups observed are biased toward protein/gene-centric curation, whereas there are other workflows, such as those for model organisms, chemicals, that may have a completely different approach to the literature curation. In fact, some model organism databases (like MGI) first do triage to select articles about the organism, then classify articles based on specific curation topics (phenotype, GO, etc). Then the conclusion from their workflow analysis could be different. Has anything come up from the survey indicating that other groups were represented in this work? For example, in the survey there is a question about database resource “Which data resource(s) do you curate?” But the result of this, which is important to learn about the databases represented, is not shown in the survey result document. I understand that the information in the survey result may be hidden to protect privacy of participants but showing the distribution of databases and/or type of databases represented (model organism vs. specific domain, like structure, PPI, etc) can shed light into bias or non-bias toward one type of curation.\n\nAnother question is about the groups observed. In the introduction, a couple of groups are mentioned that have integrated text mining pipelines in their curation work. It would be important for the study to indicate if any of the curation groups that were observed or curators surveyed used text mining tools for their work, or to include some group in observation study that do use text mining to see in what capacity and if bottlenecks in literature curation are the same.\n\nThe manuscript would be enriched by describing previous work done on this area (biocuration workflows and bottlenecks in literature curation) and compare the conclusions in this manuscript with others. There are a few papers from BioCreative that looked into this matter.1, 2, 3\n\nFinally, please consider using the term expert curation instead of manual curation. I think using the concept of manual curation in databases is not appropriate for modern times, as curators use tools to help them do all or part of their work, is not completely manual.\n\nMinor: Figures 2-5 should indicate number of participants who responded to the questions over total number of survey participants. The X-axis in Figures 3-4 need a label, same with Y-axis in Figure 5.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1622
https://f1000research.com/articles/8-1619/v1	10 Sep 19	{ "type": "Research Article", "title": "Optimal price-based control of heterogeneous thermostatically controlled loads under uncertainty using LSTM networks and genetic algorithms", "authors": [ "Taha Abdelhalim Nakabi", "Pekka Toivanen", "Pekka Toivanen" ], "abstract": "In this paper, we consider the problem of thermostatically controlled load (TCL) control through dynamic electricity prices, under partial observability of the environment and uncertainty of the control response. The problem is formulated as a Markov decision process where an agent must find a near-optimal pricing scheme using partial observations of the state and action. We propose a long-short-term memory (LSTM) network to learn the individual behaviors of TCL units. We use the aggregated information to predict the response of the TCL cluster to a pricing policy. We use this prediction model in a genetic algorithm to find the best prices in terms of profit maximization in an energy arbitrage operation. The simulation results show that the proposed method offers a profit equal to 96% of the theoretical optimal solution.", "keywords": [ "Artificial intelligence", "Artificial neural networks", "Customer behavior learning", "Demand response programs", "Energy arbitrage", "LSTM", "Partial observability", "Price elasticity of demand", "Profit maximization", "Smart grid", "thermostatically controlled loads." ], "content": "Abbreviations\n\nDR Demand Response\n\nGA Genetic Algorithm\n\nLSTM Long-Short Term Memory\n\nMDP Markov Decision Process\n\nTCL Thermostatically controlled load\n\n\nIndices\n\nn Index for TCL units n = 1,2, …,30\n\nt Index for time step, t = 1, 2, …, 24\n\n\nParameters\n\nf Transition function\n\ng State approximation function\n\nH Control horizon\n\nLmax Maximum load capacity1\n\nLmax Load threshold\n\nN Number of TCL units to control\n\nPN Population size\n\nRmax Revenue cap\n\nU Action space\n\nW Random process space\n\nX State space\n\n\nVariables\n\nC Candidate state vector in LSTM network\n\nCt Cost function at time t\n\nΔTt Gap between the outdoor and indoor temperatures [°C]\n\nh Control policy\n\nht Hidden state vector of LSTM network\n\nIn,t Input matrix of LSTM network\n\nPt Selling electricity price at time t [€ cent/kW]\n\nPt Wholesale electricity price at time t [€ cent/kW]\n\nPt,max Maximum selling price at time t [€ cent/kW]\n\nPt,min Minimum selling price at time t [€ cent/kW]\n\nPw Probability distribution\n\np Control action reward\n\nTt Temperature at time t [°C]\n\nut Control action at time t\n\nxt State at time t\n\n\nIntroduction\n\nIn a power network relying on distributed and renewable energy resources, the exploration of new sources of flexibility is a key factor for its stability. Given the intermittent nature of renewable energy resources, it is challenging to maintain the power balance under normal operating conditions in a grid with deep penetration of these resources. Therefore, more integration of renewable resources increases the need for ancillary services such as regulation reserve and load following requirements1. However, using traditional fossil fuel generators to provide these reserves will decrease the net carbon benefit from renewables, weaken generation efficiency and will be economically untenable. Alternatively, demand-side resources can play a key role in supplying the regulation service needed for deep renewable integration with zero-emission operations. Demand-side resources such as thermostatically controlled loads (TCLs), electric vehicles and strategic storage can contribute to ancillary services by acting as a source of flexibility to the grid. Unlike the traditional demand-side management programs, such as peak load shaving and emergency load management, the exploration of higher flexibility from the above-mentioned loads has a big potential in offering more lucrative and faster ancillary services. The potential of these sources of flexibility is reflected on the energy market. Electricity prices fluctuate according to the availability and demand of energy. This can open considerable opportunities for energy arbitrage2.\n\nA significant potential for provision of flexibility resides in TCLs such as air conditioners ACs, heat pumps, water heaters, and refrigerators. TCLs represent a high percentage of the total electricity consumption3,4. The nature of TCLs permits them to act as a thermal storage which makes it possible to adjust their electricity consumption while maintaining the temperature requirements and the comfort level of the end user. The idea of TCL flexibility relies on the principle that the temperature constraints specified by the users, can be fulfilled by different power trajectories. Finding the optimal trajectory that provides the required flexibility and high lucrative ancillary service is the subject of several studies5–7. However, this problem requires real-time information about the state of TCLs, their envelope temperature and their behavior in response to temperature dynamics. In most of the cases, this information is only partially available and requires qualitative or quantitative models to estimate it. It is also possible to use model-free approaches to solve the problem of uncertainty and find near-optimal power trajectories2.\n\nThe optimal power trajectory for a cluster of TCLs is then translated to individual or aggregated control signals using a variety of control methods. Control methods can be categorized into intrusive forms, including direct and indirect control, and non-intrusive form using price proxies. The direct intrusive form of control consists of directly controlling the on/off states of the TCLs, the indirect intrusive form consists of controlling the parameters of TCLs, such as the temperature set points and the switch cycles and the non-intrusive form of control uses dynamic prices to steer the consumption of TCLs relying on price-based demand response programs. The intrusive form requires an aggregator contracting with each TCL unit holder for taking control of their TCLs with the condition that their temperature constraints will be respected throughout the control period. The non-intrusive approach relies on the end user’s involvement and response to a given control signal in return of a certain incentive or special pricing. The users’ response to these signals can also be an automatic response to electricity prices throughout the day using home energy management systems or embedded TCL controllers8.\n\nIntrusive control of TCLs has a big potential in offering a wide range of flexibility and market opportunities for the aggregators. It offers a faster response to control signals and permits the design of a more reliable energy arbitrage strategy compared to non-intrusive control through price proxies. However, the implementation of the technological requirements for an intrusive control on a large scale can be challenging due to its high financial requirements. Additionally, the question of whether the consumers are ready to give up the control of their TCLs to an external party can also be a barrier for the implementation of these programs. According to 9, the integration of end users in the demand response (DR) programs is a key factor for its success. Several smart grid projects were analyzed from this perspective and the conclusions suggest that more attention should be given to the domestication of these technologies and their adaptation with the users’ experience considering their social dimensions such as individual behavior, education, and income level9,10,11. It is therefore necessary to include all these factors in the design of a DR program. Non-intrusive control, on the other hand, has fewer constraints regarding the users’ comfort and data privacy. It makes the end user feel included in the decision making of the grid and involved in the energy management. This discussion can serve as a benchmark when making the choice of the control strategy and the implementation of a large-scale DR program.\n\nIn our paper, we choose to implement a non-intrusive control using dynamic electricity prices. We first formulate the problem as a Markov decision process (MDP)12, where the policy consists of a sequence of electricity prices. The agent is assumed to have no prior knowledge or data about the state of TCL units except their real time power consumption. The idea is to use data-driven models that can learn the consumption patterns of each individual TCL unit and their response to temperatures and prices. We use a long-short-term memory (LSTM) neural network architecture to learn individual TCL units’ behaviors as in 13. This method can overcome the problem of uncertainty and the diversity of power consumption preferences in response to varying prices. The aggregator uses these models to simulate the aggregate response TCLs to different pricing schemes during a certain control horizon. An optimization algorithm is then applied to find the best pricing strategy given an objective function. When controlling a cluster of TCLs, different objective functions are considered in the literature, such as tracking a balancing signal7 or energy arbitrage5. In this work we adopt an energy arbitrage objective function, where we maximize the profit of an aggregator that buys electricity from the wholesale market and sells it in the retail market to end users with TCL units. A genetic algorithm is implemented to find the best pricing solution of the aggregate TCLs.\n\n\nRelated work and contributions\n\nThe literature contains extensive research concerning TCL control and their flexibility potential.\n\nMost early studies, as well as current work, focus on direct intrusive control methods and frameworks. Early work that tackled aggregated modeling of TCLs can be found in 14 and 15. The solution computation and controller design of these approaches is considerably difficult, which represents a drawback for these approaches. These issues were mitigated in more recent works5,7,16 using a different class of linear population-bin transition models based on Markov chains. Other approaches have proposed time-varying battery models with dissipation such as 17 or without dissipation as in 18. These approaches were used to compute near-optimal control trajectories with a reduced computational cost. Although optimal pricing for demand side management has been thoroughly studied in the literature19–21, the price-based control of TCLs remains only briefly addressed in the literature. In 22, the operating reserve capacity of aggregated heterogeneous TCLs was evaluated using a TCL model that takes into consideration consumer behavior. The price-based approach was also addressed from the consumer perspective in 23. The objective of the proposed method was mainly to find the optimal set point change in response to electricity prices in other to minimize the increases in the electricity bill due to dynamic pricing. The power gain from this control scheme was then used for load following supply. Another approach was proposed to find the equilibrium between the electricity prices and the users’ comfort. Using a Stackelberg game approach, authors in 24 presented a unique Stackelberg equilibrium that maximizes the utility function and minimizes dissatisfaction cost of TCLs users. A similar approach was proposed in 25 and 26 using a mean-field game approach to find the best pricing scheme considering TCLs as price-responsive rational agents.\n\nDeep learning and other machine learning methods are largely applied in DR programs27. The implementation of a TCL cluster control program faces the problem of uncertainty and heterogeneity of the TCL units’ behaviors in response to control prices. Consequently, many researchers were interested in using machine learning models that can learn aggregate or individual behavior of TCL units under partial observability. A model-free reinforcement learning was early proposed in 28 for TCL control that gives similar results as model predictive approaches. Reinforcement learning approaches were also used in29 to control domestic water buffers according to a local photovoltaic production for the maximization of self-consumption. More recently, the success of deep reinforcement learning approaches has inspired more researchers to tackle the problem of direct TCL control using deep reinforcement learning. Authors in 30–33 have used different deep neural architectures for automatic estimation of the TCLs’ state’s features in a batch reinforcement learning model. The same authors have later provided a comparison of the different architectures in 33,34. The LSTM architecture has outperformed the other deep neural network architectures. These works focused only on deep Q-learning, which is based on the estimation of a quality function for every potential action before performing the optimization. In 35 Deep policy gradient method was explored along with deep Q-learning for an on-line energy optimization of the buildings.\n\nFollowing the above-mentioned literature and the success of LSTM networks in mitigating the problem of partial state information and solving long-term dependency problem13,33,34, we propose a two-step pricing optimization method for the exploration of TCL flexibility in energy arbitrage. This paper addresses the need for new non-intrusive TCL control methods via electricity prices proxies, so far lacking in the scientific literature. The proposed method relies on LSTM networks learning individual TCL unit behavior and the prediction of individual responses to electricity prices. The individual predictions are aggregated to form an overall prediction model. This model is used in a genetic algorithm (GA)-based optimization algorithm to maximize a retailer’s profit considering grid and energy cost constraints. To the best of the authors’ knowledge, this is the first work that uses LSTM networks in a non-intrusive TCL control problem based on electricity prices within a DR program. The main contributions of this paper are the following:\n\nAn MDP formulation of the price control problem where the policy is the set of electricity prices during a control horizon.\n\nAn LSTM network for learning the individual behavior of TCL units in response of electricity prices and temperatures.\n\nAn aggregation of individual TCL units’ behaviors, in response to prices, to derive a global estimation of the potential response of the TCL units cluster.\n\nA genetic algorithm that uses the aggregated information from the LSTM networks to optimize the lucrative benefits from an energy arbitrage operation.\n\n\nProblem formulation\n\nWe consider a cluster of residential households powered by electricity from the same retailer or utility company. The households are equipped with smart meters and TCLs that can react to electricity prices and indoor temperatures. The retailer implements a price-based DR program that announces electricity prices for a certain time horizon in such a way that maximizes an objective function. The optimization is based on an estimated information about the responsiveness to electricity prices and temperatures. Before discussing the pricing optimization approach, we formulate the problem as an MDP12. An MDP is defined by its state space X, its action space U, and its transition function f, which defines the dynamics between the current state xt ∈ X and the next step xt+1 under a control action ut ∈U and subject to a random process w ∈ W with a probability distribution pw (., xt). The transition equation is defined as follows:\n\n\n\nThe objective of this process will be to find a policy h: X→U that minimizes or maximizes a cost function or a reward function throughout the control horizon starting from a state x1 denoted by:\n\n\n\nwhere ρ is the reward or the cost of each time step k given an action ht. Unlike the classic Q-iteration methods, the policy is characterized directly by sum of rewards during a time horizon H. The optimization is performed on the set of actions during the time horizon H and the fitness function is the cost function Rh of the policy h. For each policy h, a corresponding sequence of states is estimated implicitly by the forecasting model.\n\nThe agent is only able to measure a partial observation of the true state i.e. no information about the indoor temperatures, resulting in a partially observable Markov decision problem. The observable state space X consists of two variables: the outside temperature, and the electric load:\n\n\n\nSince the observable state space only includes part of the true state, it is not possible to directly model future state transitions. Yet this remains convenient when following the results from 13 that we can predict the next step electric load Lt+1 using the information of outdoor temperature Tt, the electric load Lt and the electricity price Pt+1. The state is extended with sequences of past observations of states and actions, which results in a non-Markovian state.\n\nFor each TCL, the electric load is approximated by:\n\n\n\nWe assume that the outside temperatures’ forecasts are available for every future timestep in the control horizon.\n\nThe control action ut consists of the electricity price that the retailer announces for each time step of the control horizon. As mentioned earlier, even though the retailer is not controlling the TCLs directly, we assume that the TCLs react directly to electricity prices. Therefore, the electricity price controls the state by influencing the amount of energy consumed during a timestep t. The next state is then defined by:\n\n\n\nAccording to the existing literature, the control of TCLs clusters can be performed considering different objective functions. For instance, the objective can be tracking a balancing signal or energy arbitrage. In this work we consider an energy arbitrage problem where a retailer is trying to maximize their profit. However, the framework and methods presented here might as well be applied to different objective functions. We consider the profit as the difference between the revenue and the cost function. We assume that the cost function Ct(Lt) is convex increasing in Lt for each timestep as formulated in 36.\n\n\n\nwhere, q > 0 is a constant, pt > 0 is the electricity price in the wholesale market and c > 0 is a fixed cost.\n\nIn order to avoid overload during peak times, we introduce a maximum load capacity of the power network, denoted Lt,max at each timestep. Therefore, we have the following constraint:\n\n\n\nThe revenue is the bill that customers would pay for using the energy during the time window H:\n\n\n\nUsually, there exists a total revenue cap, denoted as Rmax, for the retailer. Therefore, we need to add the revenue constraint to improve the acceptability of the retailer’s pricing strategies, i.e., without such a constraint, the retail prices will keep going up to a level which is against energy regulations as well as financially unacceptable to the customers. As a result, we have the following constraint:\n\n\n\nMoreover, for each timestep t ∈ H, we define the minimum and maximum price that the retailer (utility company) can offer Pt,min and Pt,max, we have:\n\n\n\nPt,min and Pt,max are usually designed based on historical prices, market competition, customers’ acceptability, and the wholesale price. It is reasonable to assume that the price the retailers can offer is greater than the wholesale price for each hour, and there exists a price cap for the retail prices due to retail market competition.\n\nFinally, the control problem defined the optimization of the price vector P, during the time horizon H, can be modeled as follows:\n\n\n\nsubject to constraints:\n\n\n\n\n\n\n\n\nMethods and implementation\n\nGiven the partial observability of this problem, the methods proposed in this paper are nondeterministic. An LSTM network is used to estimate the next states given an initial state and a pricing policy. The method consists of learning the individual behavior of each TCL agent n using an LSTM method as illustrated in 13. The N estimation models will predict the reaction Ln,t+1 of each TCL to a state x and a pricing action Pt. The overall estimated load Lt is the sum of all the load predictions as in (7). Given this estimation model, we apply a genetic algorithm to find the best pricing policy.\n\nLSTM networks are recurrent neural networks that consist of memory blocks. These memory blocks replace the summation units in the hidden layers in a standard recurrent neural network. The input vector and the hidden state vector are passed through the forget gate to determine the keeping rate of the cell state components. The same vector is passed through the input gate to determine how much of the new state candidate C can pass to the new cell state. Finally, the output gate will decide how much of the transformed state cell vector can be passed to the next hidden state vector ht. Following 13, the proposed LSTM network consists of multiple layers of LSTM cells followed by a fully connected layer as illustrated in Figure 1. In the case of our model, the input In,t is a 2 x 3 matrix that consists of the electric loads, the temperatures and the electricity prices as follows:\n\nThe model uses the information about temperatures, loads and price in the previous timesteps to predict the load L(t). Since this is a regressions problem, the fully connected layer uses a linear activation function.\n\n\n\nThe LSTM network recurrently uses the historical information of loads, temperatures and prices to predict electric load for an individual TCL n, in the next timestep. The aggregation of these predictions gives an approximation of g function mentioned in the previous section.\n\nInitially, for each TCL agent n ∈ N we train an LSTM network based on the historical reactions of these TCLs to prices and temperatures. We assume that a DR program is implemented during a long period, enough to collect a sufficient amount of data related to the reactions of TCL agents to prices and temperatures.\n\nDue to the discontinuous nature of the objective function and the complicated dependency between the function electric load L and the electricity prices P, the conventional nonlinear optimization methods are not usable for this problem. Therefore, GA-based optimization algorithms are more suited for this problem37. The proposed GA algorithm uses rank selection and value encoding38. Each chromosome represents a pricing policy P and consists of a vector of size H. We use uniform crossover39 and non-uniform mutation40. The constraints are handled by the approach proposed in 41.\n\nThe proposed GA-based optimization algorithms for TCL pricing control are given in Algorithm 1 and Algorithm 2.\n\n1: Population Initialization, i.e., generating a population of PN chromosomes randomly; each chromosome denotes a pricing policy for the next time horizon H.\n\n2: for i=1 to PN do\n\n3: Concatenate the price vector to the temperature forecasts of the next time horizon.\n\n4: for each TCL agent n in N do:\n\n5: Use LSTM network iteratively to predict (Ln,t)t∈H using Algorithm 2.\n\n6: end for\n\n7: Calculate Lt, Ct(Lt) ∀t ∈ H, and R\n\n8: Check the feasibility of policy P regarding the constraints. Handle the invalid individuals by the approach proposed in []. Then calculate the fitness value of policy P.\n\n9: end for\n\n10: Create a new generation of chromosomes by using the selection, crossover, and mutation operations of the GA.\n\n11: Repeat steps 2–11 until the stopping condition is reached.\n\n12: Announce the best price vector via the two-way communication infrastructure at the beginning of the control horizon.\n\n1: Build the initial input matrix In,0 using the initial values of prices, loads and temperatures.\n\n2: for t=0 to H do\n\n3: Use the input matrix In,t to predict Ln,t+1\n\n4: Concatenate L, T and P with the last line of the input matrix In,t to build the next input matrix:\n\n\n\n5: end for\n\n6: return (Ln,t)t∈H\n\nIn Algorithm 1, we initialize a population of NP pricing policies at step 1. For each policy P we perform steps 2–6 to evaluate the fitness function and the feasibility for each policy. The evaluation of policies is performed using LSTM sequence prediction presented in Algorithm 2. The best policies are selected, and a new generation is created using crossover and mutation operations in step 10. This process is repeated until a stopping condition or maximum number of iterations is reached. At the end of the optimization process, the best pricing policy is selected, and prices are announced to TCL agents via two-way communications technology. After each control episode, the LSTM learning models are updated according to the new data collected from the actual response to the implemented electricity prices.\n\n\nResults\n\nIn this section we evaluate the functionality of the proposed pricing control methods. A set of numerical experiments were performed on a simulation scenario comprising a population of 30 TCLs exposed to dynamic electricity prices during a period where the outdoor temperatures change significantly. The thermal inertia of each TCL allows the electric demand to be shifted towards lower price moments. The TCL agents determine the amount of electricity to be consumed at each timestep according to the indoor temperature and the electricity prices. The objective of TCL agents is to maintain a reasonable comfort level while minimizing the electricity bill. Therefore, the different TCL agents have different reactions given a set of prices and temperatures depending on individual user’s preferences and buildings’ characteristics. We define a control timestep of 1 hour and a control horizon of 6 hours. The choice of the control horizon is justified by the limited ability of LSTM to predict large sequences of the future electric loads. The control horizon is chosen in a way that minimizes the number of times the retailer runs the control algorithms and announces the prices, while keeping a good accuracy of the LSTM predictions.\n\nFollowing 13 the simulation data is generated using two fuzzy logic systems with the following assumptions:\n\nThe TCL agents are reacting to indoor temperatures and electricity prices.\n\nThe difference between the outdoor and indoor temperature ∆T depends on the building characteristics and the amount of energy spent in heating/cooling in previous timesteps.\n\nTCL agents are operating during the day to maintain a comfortable temperature of the space while taking into consideration the electricity price in a given hour. Fuzzy logic is used in this problem because it can model non-qualitative concepts like “hot temperature” or “low price”. The combination of the two fuzzy logic systems delivers the load Ln,t+1 using the outdoor temperature Tt and the electricity price Pt+1. The simulation is performed with different parameters to generate diverse data for 30 TCL agents. The temperature and price data used for the simulation are taken respectively from the Kaisaniemi observation station in Helsinki, available online in 42, and Elspot DA electricity prices in Finland43 for the period between 1st January 2017 and the 7th September 2018. The generated dataset consists of 14,734 data points for each TCL agent.\n\nThe data generated from the above-mentioned simulations is used to train the LSTM networks to learn the behavior of each individual TCL agent. The hyperparameters and structure of the LSTM networks are chosen according to the results of 13 and summarized in Table 1.\n\nThe results are evaluated using validation data generated from the same simulations. Figure 2a illustrates the learning results for three TCL agents during different time periods with different temperatures and prices. Figure 2b illustrates the comparison between the real and predicted average power consumption of the 30 TCL agents cluster. The power curves show that the TCL agents’ responses to prices and temperatures are slightly different. In general, the power consumption is high when the temperatures and electricity prices are low and vice-versa. The comparison between the true load curves and the predicted load curves show a very small prediction error per hour in most cases. The true and predicted load curves have similar shapes and significant resemblances. The peaks and valleys are also predicted accurately in most of the cases, which gives a valuable insight for demand side management.\n\n(a) Power consumption of different TCL agents in response to electricity prices and outdoor temperatures. (b) Average real and predicted power consumption of the cluster surrounded by an envelope containing 9% of the power consumption profiles for different days.\n\nWe run the GA optimization algorithm on a population of size 100 for 100 iterations. The parameters used for the optimization are summarized in Table 2. The optimization process is graphically presented in Figure 3. The learning process is measured by the fitness of the best individual in the population at each iteration. Figure 4 illustrates the results of the best pricing solutions for one day. Figure 4a is an illustration of the electricity prices fluctuations during the 24 hours. Figure 4b shows a comparison between the power consumption of the whole cluster under original prices and the power consumption under optimized prices. Figure 4c presents the revenue and profit that the retailer would make under original and optimized prices. Figure 4d presents daily bill of each user of the cluster under original and optimized prices.\n\n(a) Optimized prices solution for 24 hours. (b) Revenue and profit under original and optimized prices for 24 hours. (c) Total electricity consumption under original and optimized prices. (d) Daily electricity bills under original and optimized prices.\n\nThe results show a general increase in prices throughout the day. However, this increase didn’t result in an increase in the daily electricity bills. Most of customers will be paying a slightly lower amount per day. This is a consequence of upper limit constraint on the revenue described in (12). The overall consumption of electricity was decreased comparing to the original pricing scheme which gives a good idea about the potential energy saving that an optimal pricing strategy can offer.\n\nIn order to validate the performance of the proposed algorithm, we consider a case where we have a full access to TCL units’ behavior, i.e. the exact electricity consumption of each TCL unit given temperatures and prices at each timestep. The optimization is performed with direct access to the simulation model described above, which provides full observability and perfect information about the TCLs. This theoretical setup can serve as a benchmark of our method. It can be seen as an upper limit on the profit possibly made by the aggregator without violating the constraints.\n\nThe results illustrated in Figure 5a–d, show that the proposed methods have performed very similarly to the benchmark. The hourly prices in Figure 5a, are only slightly shifted from the benchmark prices during most of the day. The difference is only significant in 2 to 3 points. The same observation can be made for the revenues and profits in Figure 5b and electricity consumption in Figure 5c. The comparison of daily bills under optimized prices and benchmark prices in Figure 5d shows a slight rise in the electricity bill in the benchmark model for most customers. This can be explained by the slight increase in prices illustrated in Figure 5a.\n\n(a) Comparison between benchmark and optimized prices. (b) Hourly revenues and profits under optimized prices and benchmark prices. (c) Hourly total electricity consumption under optimized prices and benchmark prices. (d) Daily electricity bills under optimized and benchmark prices.\n\nThe daily revenues and profits under original, optimized and benchmark prices are compared in Figure 6. The comparison shows a closely similar revenue in the three cases. The optimized prices have given a slightly smaller revenue compared to the revenue from original and benchmark prices. However, the profit from original prices is considerably smaller than the profit from optimized prices. The latter is only slightly smaller than the benchmark’s profit. Numerically, the profit from the proposed methods is 95.97% of the optimal benchmark profit. This observation shows that an increase in the profit can be made without an increase in the revenue when the prices are optimized correctly.\n\n\nDiscussion and conclusion\n\nIn this paper, we demonstrated the effectiveness of a new TCL control using electricity price proxies. The control policy consists of a sequence of prices influencing the electricity consumption from TCLs. The problem was formulated as a Markov decision process with non-Markovian state to handle the sparse observations of the TCL cluster’s state. We extend the observable state with sequences of past observations to approximate the transition function using an LSTM architecture. The LSTM network is used to capture the individual behavior of TCLs under price-based DR. The individual models are aggregated to approximate the next state of the cluster. This approximation is used iteratively in a genetic algorithm to evaluate the potential profit from an energy arbitrage operation and find the optimal pricing policy for a given control horizon. The LSTM models are updated every 24 hours to capture the changes in the TCL units’ behavior.\n\nThe experiment consists of a retailer agent buying electricity from the wholesale market and selling it to a group of residential TCLs. The agent can only measure the electricity consumption of each TCL and the outside temperature. The agent has access to a significant amount of historical data from an already implemented DR program. Which allows it to train the LSTM models for each TCL unit and perform an optimization on the electricity prices.\n\nWe first evaluate the performance of the LSTM network by comparing the real and predicted loads from 30 TCL units during different days. The predicted load profiles are closely similar to the real load profiles both at individual and aggregate level. The optimization relies on a genetic algorithm with a profit maximization objective. The results of the optimization show that the proposed methods offer a much higher daily profit than the original prices and 95.97% of the optimal profit from a model that has full observation of the state.\n\nThe flexibility offered by TCLs is a high potential for ancillary services required for a deep integration of renewable energy sources in the grid. An energy arbitrage operation can offer a service to the grid by exploiting this flexibility using direct or indirect control. The partially observable state and the uncertainty of the TCL response to prices was tackled in this paper with an LSTM network using past observations and actions. The LSTM network offered a high performance by extracting relevant features of the hidden state using its internal memory cell, allowing it to process sequences of sparse observations to learn the hidden patterns of power consumption.\n\n\nData availability\n\nFigshare: LSTM+GA data, https://doi.org/10.6084/m9.figshare.9746786.v144.\n\nThis project contains the following underlying data:\n\nData used by the fuzzy logic simulation model such as temp_prices and temperatures.\n\nData generated by the fuzzy simulator such as fuzzy_outxx.csv and used to train the LSTM models.\n\nData related to the optimization process such as results and GA_pricing, optimized_prices_loads\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).\n\n\nSoftware availability\n\nCode for analysis available from: https://github.com/tahanakabi/Optimal-Price-Based-control-of-heterogeneous-thermostatically-controlled-loads-under-uncertainty-usi\n\nArchived code as at time of publication: http://doi.org/10.5281/zenodo.338361545\n\nLicense: MIT", "appendix": "Footnotes\n\n1This work was supported by The Jenny and Antti Wihuri Foundation, FINLAND.\n\n\nReferences\n\nHalamay DA, Brekken TKA, Simmons A, et al.: Reserve requirement impacts of large-scale integration of wind, solar, and ocean wave power generation. 2010. Publisher Full Text\n\nMathieu JL, Kamgarpour M, Lygeros J, et al.: Arbitraging Intraday Wholesale Energy Market Prices With Aggregations of Thermostatic Loads. IEEE Transactions on Power Systems. 2015; 30(2): 763–772. Publisher Full Text\n\nD&R International, Ltd.: 2011 Buildings Energy Data Book. 2012. Reference Source\n\nU. E. I. Administration: U.S. Energy Information Administration. 2010. Reference Source\n\nMathieu JL, Kamgarpour M, Lygeros J, et al.: Energy arbitrage with thermostatically controlled loads. 2013. [Accessed 11 6 2019]. Publisher Full Text\n\nMaasoumy M, Razmara M, Shahbakhti M, et al.: Selecting building predictive control based on model uncertainty. 2014. [Accessed 11 6 2019]. Publisher Full Text\n\nKoch S, Mathieu JL, Callaway DS: Modeling and Control of Aggregated Heterogeneous Thermostatically Controlled Loads for Ancillary Services. 2011. [Accessed 11 6 2019]. Reference Source\n\nSaha A, Kuzlu M, Pipattanasomporn M, et al.: Enabling Residential Demand Response Applications with a ZigBee-Based Load Controller System. 2016; 2(4): 303–318. [Accessed 11 6 2019]. Publisher Full Text\n\nVerbong GPJ, Beemsterboer S, Sengers F: Smart grids or smart users? Involving users in developing a low carbon electricity economy. Energy Policy. 2013; 52: 117–125. Publisher Full Text\n\nYan X, Ozturk Y, Hu Z, et al.: A review on price-driven residential demand response. Renew Sust Energ Rev. 2018; 96: 411–419. Publisher Full Text\n\nHansen M, Borup M: Smart grids and households: how are household consumers represented in experimental projects? Tech Anal Strat Manag. 2018; 30(3): 255–267. Publisher Full Text\n\nLittman ML: Markov Decision Processes. 2001. 9240–9242. [Accessed 12 6 2019]. Publisher Full Text\n\nNakabi TA, Toivanen P: An ANN-based model for learning individual customer behavior in response to electricity prices. Sustainable Energy, Grids and Networks. 2019; 18. Publisher Full Text\n\nIhara S, Schweppe FC: Physically based modeling of cold load pickup. IEEE Transactions on Power Apparatus and Systems. 1981; 100(9): 4142–4150. Publisher Full Text\n\nMalhame R, Chong CY: Electric load model synthesis by diffusion approximation of a high-order hybrid state stochastic system. IEEE TRANSACTIONS ON AUTOMATIC CONTROL. 1985; 30(9): 854–860. Publisher Full Text\n\nMathieu JL, Koch S, Callaway DS: State estimation and control of electric loads to manage real-time energy imbalance. IEEE Trans Power Syst. 2013; 28(1): 430–440. Publisher Full Text\n\nHao H, Sanandaji BM, Poolla K, et al.: Aggregate Flexibility of Thermostatically Controlled Loads. IEEE Transactions on Power Systems. 2015; 30(1): 189–198. Publisher Full Text\n\nKamgarpour M, Ellen C, Soudjani SEZ, et al.: Modeling options for demand side participation of thermostatically controlled loads. In 2013 IREP Symposium Bulk Power System Dynamics and Control-IX Optimization, Security and Control of the Emerging Power Grid,. Rethymno, Greece, 2013. Publisher Full Text\n\nMeng FL, Zeng XJ: A Profit Maximization Approach to Demand Response Management with Customers Behavior Learning in Smart Grid. IEEE Trans Smart Grid. 2016; 7(3): 1516–1529. Publisher Full Text\n\nJia L, Zhao Q, Tong L: Retail pricing for stochastic demand with unknown parameters: An online machine learning approach. In 2013 51st Annual Allerton Conference on Communication, Control, and Computing (Allerton). Monticello, IL USA, 2013. Publisher Full Text\n\nDehghanpour K, Nehrir HM, Sheppard JW, et al.: Agent-Based Modeling of Retail Electrical Energy Markets With Demand Response. IEEE Transactions on Smart Grid. 2018; 9(4): 3465–3475. Publisher Full Text\n\nXie D, Hui H, Ding Y, et al.: Operating reserve capacity evaluation of aggregated heterogeneous TCLs with price signals. Applied Energy. 2018; 216: 338–347. Publisher Full Text\n\nJay D, Swarup K: Price Based Demand Response of Aggregated Thermostatically Controlled Loads For Load Frequency Control. In 17TH NATIONAL POWER SYSTEMS CONFERENCE. 2012. Reference Source\n\nWang P, Zou S, Wang X, et al.: A Stackelberg Game Approach for Price Response Coordination of Thermostatically Controlled Loads. Applied Sciences. 2018; 8(8): 1370. Publisher Full Text\n\nDe Paola A, Trovato V, Angeli D, et al.: A Mean Field Game Approach for Distributed Control of Thermostatic Loads Acting in Simultaneous Energy-Frequency Response Markets. IEEE Transactions on Smart Grid. Early Access 2019; 1–1. Publisher Full Text\n\nGrammatico S, Gentile B, Parise F: A Mean Field control approach for demand side management of large populations of Thermostatically Controlled Loads. In 2015 European Control Conference (ECC). Linz, Austria, 2015. Publisher Full Text\n\nNakabi TA, Haataja K, Toivanen P: Computational Intelligence for Demand Side Management and Demand Response Programs in Smart Grids. In 8th International conference on bioinspired optimization methods and their applications. Paris, 2018. Reference Source\n\nCan Kara E, Berges M, Krogh B, et al.: Using smart devices for system-level management and control in the smart grid: A reinforcement learning framework. In 2012 IEEE Third International Conference on Smart Grid Communications (SmartGridComm). Tainan, Taiwan, 2012. Publisher Full Text\n\nDe Somer O, Soares A, Kuijpers T, et al.: Using Reinforcement Learning for Demand Response of Domestic Hot Water Buffers: a Real-Life Demonstration. Cornell University, 2017. Reference Source\n\nRuelens F, Claessens BJ, Quaiyum S, et al.: Reinforcement Learning Applied to an Electric Water Heater: From Theory to Practice. IEEE TRANSACTIONS ON SMART GRID. 2018; 9(4): 3792–3800. Publisher Full Text\n\nClaessens BJ, Vrancx P, Ruelens F: Convolutional Neural Networks for Automatic State-Time Feature Extraction in Reinforcement Learning Applied to Residential Load Control. IEEE Transactions on Smart Grid. 2018; 9(4): 3259–3269. Publisher Full Text\n\nRuelens F, Claessens BJ, Vandael S, et al.: Residential Demand Response of Thermostatically Controlled Loads Using Batch Reinforcement Learning. IEEE Transactions on Smart Grid. 2017; 8(5): 2149–2159. Publisher Full Text\n\nRuelens F, Claessens BJ, Vrancx P, et al.: Direct Load Control of Thermostatically Controlled Loads Based on Sparse Observations Using Deep Reinforcement Learning. Cornell University. 2017. Reference Source\n\nPatyn C, Ruelens F, Deconinck G: Comparing neural architectures for demand response through model-free reinforcement learning for heat pump control. In: 2018 IEEE International Energy Conference (ENERGYCON). Limassol, Cyprus, 2018. Publisher Full Text\n\nMocanu E, Constantin Mocanu D, Nguyen PH: On-line Building Energy Optimization using Deep Reinforcement Learning. IEEE Transactions on Smart Grid. 2018; 10(4): 3698–3708. Publisher Full Text\n\nMohsenian-Rad AH, Wong VW, Jatskevich J, et al.: Autonomous Demand-Side Management Based on Game-Theoretic Energy Consumption Scheduling for the Future Smart Grid. IEEE Transactions on Smart Grid. 2010; 1(3): 320–331. Publisher Full Text\n\nHolland JH: Genetic Algorithms. Scientific American. 1992; 267(1): 66–72. Publisher Full Text\n\nBlickle T, Thiele L: A comparison of selection schemes used in evolutionary algorithms. Evol Comput. 1996; 4(4): 361–394. Publisher Full Text\n\nSyswerda G: Uniform crossover in genetic algorithms. Proceedings of the 3rd International Conference on Genetic Algorithms. San Francisco, CA USA, 1989. Reference Source\n\nNeubauer A: Adaptive non-uniform mutation for genetic algorithms. In: Computational Intelligence Theory and Applications. Berlin, Heidelberg, 1997; 24–34. Publisher Full Text\n\nDeb K: An efficient constraint handling method for genetic algorithms. Comput Methods Appl Mech Eng. 2000; 186(2–4): 311–338. Publisher Full Text\n\nWeather observations, Kaisaniemi observation station Helsinki. Finnish meteorological institute. [Accessed 8 September 2019]. Reference Source\n\nNord Pool, Elspot Day-ahead, Prices. [Accessed 8 September 2018]. Reference Source\n\nNakabi T: LSTM+GA data. figshare. Dataset. 2019. http://dx.doi.org/10.6084/m9.figshare.9746786.v1\n\nNakabi TA: tahanakabi/Deep-Reinforcenment-learning-for-TCL-control: First release (Version V1.0.0). Zenodo. 2019. http://dx.doi.org/10.5281/zenodo.3383615" }	[ { "id": "68829", "date": "25 Aug 2020", "name": "Matsukawa Shun", "expertise": [ "Reviewer Expertise smart grid", "baseline load estimation", "machine learning", "neural networks", "time-series", "LSTM" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe sequence length of the LSTM model is so short that it does not show superior predictive ability compared to other models.\n\nI couldn't understand the significance of Figure 2 because of its lower resolution. This issue needs to be resolved.\n\nOn my own interpretation of Figure 2, I felt that the orange breakline showing the LSTM prediction results failed to learn the response of the TCL agent because it was heavily dependent on the load of the previous step.\n\nIn Figure 4 and 5, the optimized loads on time 14-15 are changes suddenly. I felt it is necessary to investigate the cause of it.\nComment) I felt that the real key to this study was not the optimization, but the accuracy of TCL response prediction. Therefore, a more detailed analysis of the LSTM model would further enhance the value of this paper.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "68823", "date": "01 Sep 2020", "name": "Chiou-Jye Huang", "expertise": [ "Reviewer Expertise Big data analysis", "machine learning and deep learning applications on the Internet of Energy (IoE) and environmental science", "especially in renewable energy", "as well as electricity load demand", "electricity prices", "solar radiance", "photovoltaic power", "and PM2.5 forecasting", "and photovoltaic power plants planning design", "and operation maintenance management." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors proposed a long-short-term memory (LSTM) network to learn the individual behaviors of TCL units. The authors use the aggregated information to predict the response of the TCL cluster to the pricing policy. The authors use this prediction model in a genetic algorithm to find the best prices in terms of profit maximization in an energy arbitrage operation. The simulation results show that the proposed method offers a profit equal to 96% of the theoretically optimal solution. I recommend minor revisions. I recommend the following revisions. In addition, there are some questions that need to be explained below:\nEnglish language should be carefully checked and carefully check paper for language typos.\n\nSome figures are not needed.\n\nAll the figures are unclear and hard to read, please update to a clear version.\n\nThe authors must provide a detailed flowchart of the methodology of the paper.\n\nThe conclusion section is missing some perspective related to future research work.\n\nReferences are too few and must be updated in recent years. I suggest authors should add related references.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "68826", "date": "01 Sep 2020", "name": "Shanti Swarup", "expertise": [ "Reviewer Expertise Demand Response and Management" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe idea behind TCL is very good and how it effects the price.\n\nHowever, the paper is not properly presented.\n\nWhat is LSTM? long-short-term memory is contradicting what is long -short. Either it has to be long term memory (LTM) or short-term Memory (STM).\n\nDiscussions and conclusions should be separate. Difficult to identify the conclusions. It should be results and discussions.\n\nThere are three different tools employed as shown below. Why is there a need to use all these tools (DL, LTSM, GA) Deep learning is used for control of TCL loads LSTM networks for state estimation Genetic algorithms for price optimization Only one tool can be used. In fact; prediction of load for TCL is missing.\n\nWhy is a need for price optimization?\n\nThe price (LMP) is dependent on the intersection between the generation and demand. This price keeps on varying.\n\nThe social benefit or social welfare needs to be optimized and not the price. Eqn 6 is not correct.\n\nFigs 2a and 2b do not provide sufficient information to infer the contribution. The need for so many plots is questionable.\n\nOnly important results should be provided.\n\nIn-spite of the good idea and motivation, the approach used seems to be not proper.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [] } ]	1	https://f1000research.com/articles/8-1619
https://f1000research.com/articles/8-1616/v1	10 Sep 19	{ "type": "Research Article", "title": "Effectiveness of the Hospital Learning Center (Queen Sirikit National Institute of Child Health): Satisfaction with service and parents’ attitudes towards children’s illness", "authors": [ "Adidsuda Fuengfoo", "Sija Leelathanaporn", "Thanyaporn Mekrungcharas", "Kim Sakulnoom", "Sumitra Owjinda", "Piyanat Noipong", "Suphawan Srinuan", "Sarunya Kumjaroen", "Natchanan Phonok", "Adidsuda Fuengfoo" ], "abstract": "Background: All children, whether healthy or ill, should have access to equal educational opportunities. Healthcare institutions and hospitals have been approved to work with schools to establish learning centers to provide education to sick children. This study has been conducted to develop a practical model for learning centers in hospitals across Thailand to create equality and ensure valuable human resources for the future. The main goal of this study was to evaluate the effectiveness of a hospital learning center for continuing education of child patients and to determine the factors that are most appropriate study plans, the parents’ attitude about their child’s illness, and the children’s satisfaction with the learning center. Methods: The total sample population was 400, consisting of 200 parents and 200 child patients aged 4 to 18 years. The respondents were given a questionnaire to obtain their feedback using a Likert scale. Results: The most common child patients were those with chronic illness followed by those with common illnesses, and lastly children with developmental problems. All 200 children received continuing education; 20 child patients (10%) who had been evaluated received a modified education plan. After analyzing the results of satisfaction with the learning center, the scores ranged from 4.21 to 5.00 (mean = 4.28, SD = 0.62).\n\nConclusions: Sick children can continue their education at the hospital learning center in Queen Sirikit National Institute of Child Health. Study plans can be modified to suit children with chronic illnesses and developmental problems, children in primary and secondary school, and those requiring prolonged hospitalization. Parents in the study had appropriate attitudes about the disease and education of their children. Sick children gave the highest ratings showing extremely high satisfaction with the hospital learning center.", "keywords": [ "Children with illness", "Education service", "School Engagement" ], "content": "Introduction\n\nInformation technology, social networking, and the creation of a global economy continue to develop rapidly, but education is still the primary basis of child development and growth. All children, whether sick or healthy should receive equal education. The number of children living with chronic conditions has steadily risen as improvements1 in medication and pediatric care for chronic illness have resulted in more children with previously lethal conditions now surviving into adulthood and beyond1,2. The prevalence of children with chronic disease varies depending on the definition used. One survey found that the prevalence of chronic health condition in children was 13-27 percent3. The number of children in Thailand with chronic disease is not clearly defined because chronic diseases can develop slowly over time4,5; some are incurable, requiring long-term treatment with adverse effects on the patient’s and family’s daily activities5,6. Generally, children with chronic conditions have lengthy hospital stays and require continued follow-up after going home, causing major changes in the lives of patients and families. Some patients do well in home care but their condition may be too serious to permit them to return to school. Some lose a lot of school days, which prevents them from sharing social experiences7,8. The Convention on the Rights of the Child recognized education as a legal right of every child on the basis of equal opportunity9. As a member of the United Nations, Thailand has adopted a law that states “education is not limited the classroom”. Hospital and health care institutions have been approved to establish learning centers in the hospital to provide education for ill children under collaboration between hospitals and schools. Queen Sirikit National Institute of Child Health (QSNICH) is working under a project called “The Information Technology Project under the Initiative of Her Royal Highness Princess Maha Chakri Sirinhorn”, by using technology to support learning with the help of multidisciplinary team, which increases the opportunity to connect with education and society.\n\nThe Learning Center in the Hospital at QSNICH has been operating for 20 years. A previous study10 identified educational and social perspectives of child patients that influenced psychological conditions and adaptation to illness but did not address the effectiveness of the learning center. With collaboration between various sectors at the ministry level, this study has been conducted to determine the effectiveness of the center and to encourage continuing and appropriate education plans for child patients, as well as to identify the factors that relate to modification of education plans. The results of this study will be useful for developing education models of hospital learning centers for implementation at hospitals across Thailand to provide educational equality and valuable human resources for the future.\n\n\nObjectives\n\nThe primary objective of this study was to evaluate the effectiveness of the hospital learning center and to determine appropriate plans for continuing and improving the education of child patients.\n\nThe secondary objective was to identify the factors involved in the adjustment of educational plans, assess the parents’ attitudes about their child’s illness, and rate the satisfaction of the children with the learning center.\n\n\nMethods\n\nA cross-sectional survey was conducted to obtain information about the parents and the children who entered the learning center at QSNICH, Bangkok, Thailand between January 1 and December 31, 2018.\n\nThe total population was 400 persons, consisting of 200 parents and 200 sick children from the learning center in hospital at the QSNICH from January 1, 2018 to December 31, 2018. The sample size was calculated based on a previously described method11 by reviewing the number of children who returned to school (80% of all children), given p = 0.8, N = 1000. The computed result gave a sample size of 198 which led to the number of 200 children and 200 parents, giving 400 persons in total. The inclusion criteria were sick children at the learning center aged between 4 and 18 years, sick children who visited the learning center more than twice and sick children with critical illnesses at the hospital or home who still having continual treatment. The exclusion criterion was sick children at terminal stage of life or having critical illnesses. Informed consent for study participation was obtained from child patients and parents.\n\nResearchers used systematic random sampling in which the children on the name list at the learning center, QSNICH were selected in order. If any respondents refused to answer the questions, the latter name would be selected. After giving consent, the researcher started interviewing parents and sick children. For children from 4 to 7 years of age, parents were asked to join the interview12. The answers were noted, checked, and recorded for data analysis.\n\nThe questionnaire was divided into four parts; demographic data of parents, relationship between parents and children, children’s personal information and satisfaction with service at learning center using a Likert scale. The rating scores were divided into five levels. The questionnaire used, alongside an English translation, is available as Extended data13.\n\nThe variables associated in this research included parents and children’s relationship, educational level, parent’s occupation, family income, home town, severity of diseases, period of hospitalization, and satisfaction with service at the learning center.\n\nLevels of significance based on satisfaction scores of the learning center at the hospital and were divided into five levels: Strongly agree (4.21–5.00 points), agree (3.41–4.20 points), uncertain (2.61–3.40), disagree (1.81–2.60), strongly disagree (1.00–1.80). P < 0.05 was considered to indicate statistical significance.\n\nDataFax was used for data management. Descriptive statistics, including percent, mean and standard deviation were used for analyzing population data and SAS 9.4 software was used to calculate frequency, percentage, mean and standard deviation to measure attitude and satisfaction of parents and children at the center. Data on satisfaction was divided into 5 levels: strongly agree, uncertain, disagree, strongly disagree based on statistical test results. Chi-square test and t-test were used to assess variables associated with continuation of education at the learning center.\n\nThis study was been approved by the Office for Ethics in Human Research, Queen Sirikit National Institute of Child Health in 2018 (ref. REC.010/2562), and was conducted in accordance with The Declaration of Helsinki\n\n\nResults\n\nResponses to each question of the survey from each participant is available as Underlying data14.\n\nThe study categorized the children into three groups: those with common illness, those with chronic conditions and disabled children. The other population studied was the 200 caregivers, the largest percentage of which was the biological parents of the children at 87.9%, 91.5% and 93.3% for the three groups, respectively; caregivers who are relatives were caregivers for 12.1%, 8.5% and 6.7%, respectively. Neither foster parents nor teachers were guardians. The majority of caregivers were female (90%, 90% and 53% for those with common illness, those with chronic conditions and disabled children, respectively), of Thai nationality (100%, 100%, and 93%, respectively) and Buddhist religion (97%, 100%, and 93%, respectively); the non-Thai caregivers were from Myanmar and other religions included Islam and Christianity. Most parents were 31 to 40 years of age (46.1%, 44.7%, and 53.3% for those with common illness, those with chronic conditions and disabled children, respectively). Those 41 to 50 years old constituted 23%, 39%, and 40%, respectively, while parents who were 21 to 30 years old were involved with only 21%, 9%, and 0% of the children, respectively. Most parents had a bachelor’s degree or vocational education, but a small percentage went to secondary school. Occupations of most parents were in sales and agriculture or laborers. Most were married, separated, or divorced. The majority had income in the range of THB 10,000-30,000, with the rest earning THB 5,000-10,000, and only a small proportion receiving more than THB 30,000 or less than THB 5,000. The study families in general were home owners, while some rented a house or apartment and lived in the central part of Thailand. Only a few lived with relatives.\n\nAccording to Table 1, for the 200 child patients, 91 caregivers (45.5%) were parents of children with common illnesses, 94 (47%) were parents of children with chronic illnesses and 15 (7.5%) were parents of children with disabilities. In general, parents of the three categories of sick children rated each question similarly. Regarding the severity of the illnesses, 12.0% rated them at the minimum level, 46.5% at the moderate level, and 39.5% at the high level. Whenever the patients get ill, 74.5% of the caregivers always reported it to school authorities while 20.5% frequently did so. About 34.0% were very confident that their children got along well with their peers, 31.0% were moderately confident and 29% highly confident. Similarly, 44.5% of caregivers attempted to help their children with school work at a moderate level, while 23.5% had tried at a higher level. Overall, 34.5% often explained the classwork to their children and 33.5% did it all the time. About 11% were moderately confident that the school arranged good study plans for their children, 47% were moderately confident and 38.5% were highly confident. Overall, 44% of the child patients talked about what happened in their school and 18.5% did it all the time. Regarding the caregivers’ confidence to help their children to control their emotions, about 42% were very confident, and about 17% were moderately confident and 34% were extremely confident. In total, 91.37% of caregivers took their children to school on a daily basis.\n\nAs shown in Table 2, 23.1% of children with common illness were in kindergarten, 64.8% in primary school and 10.9% in secondary school before entering the learning center. After discharge, 20.8% were in kindergarten, 67.0% were in primary school and 10.9% were in secondary school. The distances from home to school before and after attending the learning center did not differ. Approximately, 16.5% lived less than 1 km away, 60.4% lived 1–5 km away and 23.1% lived the same distance as before entering the learning center. The data also showed that 29.7% went to school by bicycle or motorcycle, 39.6% by personal car, and 28.6% by public transportation; the mean of transportation was not much different from before and after entering learning center.\n\nFor patients with chronic illnesses, 10.6% were in kindergarten, 68.1% were in primary school and 21.3% were in secondary school; none of them attended vocational education. After discharge, 10.6% adopted a modified study plan for non-formal education; 6.4% of parents lived less than 1 km away while 56.4% lived 1–5 km away from school and 41.5% were the same distance before and after attending the center. With regard to transportation, 37.2% went to school by bicycle or motorcycle, 25.5% by personal vehicle, and 33.0% by public transportation; the transportation data were almost the same before and after attending the center.\n\nFor children with disabilities, while being treated and attending the center, 13.3% were in kindergarten, 66.7% in primary school, 13.3% in secondary school, and 6.7% attended an individualized education program (IEP); after discharge, 13.3% were in kindergarten, 6.7% were in primary school and 6.7% in secondary school, 73.3% attended an IEP. After discharge, none of the parents live closer to the center than 1 km, 53.3% lived 1–5 km away and 46.7% live more than 5 km away from the center. The distance was almost the same before and after being attending school at the center. For transportation, 33.3% go to school by bicycle or motorcycle, 46.7% by personal vehicle, and 20.0% by public transportation; the transportation percentages are almost the same before and after going to the center.\n\nAs shown in Table 3, the factors relating to modifications of the education plan of the school were subjected to logistic regression. Nearly 100% of the children as general patients, chronic patients, and patients with developmental problems continued their education. Patients with common illness did not adjust their education plan, although 10.6% of chronic patients and 73.3% of special patients did modify their plan. The number of chronic patients who modified their education plan was significantly greater than the number of general patients (OR = 0.05; 95% CI: 0.02-0.14). In terms of education level, 95.5% of primary school students and 53.2% of secondary school students had their education plan modified. The number of primary school students who modified their education plan was significantly higher than those at kindergarten level (OR = 2.08; 95% CI: 0.95-4.55).\n\nNotes: Percentages are shown in a row. [1] The p value was based on logistic regression. [2] No 95%CI was considered. *significant difference (p < 0.05)\n\nAs recorded in Table 4, parents’ opinions about each question across all types of illnesses are different. Considering the topic of learning, 74.7% of general patients, 69.2% of chronic patients, and 66.7% of disabled patients stated that the subjects at the center were interesting while 24.2% of general patients, 30.8% of chronic patients, and 33.3% of disabilities patients strongly agreed with this. More than half of the children enjoyed learning new things at the center (75.8% of patients with common illnesses, 70.2% of chronically ill patients and 73.3% of disabled patients). In terms of patients overall, 26.5% (23.1%, 29.8%, and 26.7% for those with common illness, those with chronic conditions and disabled children, respectively) strongly agreed that that learning at the center was enjoyable. In terms of looking forward to receiving services from the center, about 69.0% of all patients (73.6% of general patients, 65.9% of chronic patients, and 60.0% of patients with disabilities) agreed with this issue; and 22.2% of all patients (22.0% of general patients, 20.0% of chronic patients, and 22.0% of disabled patients) strongly agreed with this issue. Regarding the proper facilities supporting learning environment at the center, about 69.5% of all patients (71.4% of general patients, 67.0% of chronic patients, and 73.3% of special patients) agreed with this issue; and 31.9% of all patients (28.3% of general patients, 34.2% of chronic patients, and 35.1% of special patients) strongly agreed with this issue. In terms of advancement, varieties and adequacy of the study plan of the center, about 60.9% of all patients (65.6% of general patients, 57.7% of chronic patients, and 59.2% of special patients) agreed with this issue; and 30.0% of all patients (27.5% of general patients, 33.0% of chronic patients, and 26.7% of special patients) strongly agreed with this issue. The students also reported a variety of learning materials used in the classroom: 70.5% of patients used textbooks, (63.4% of general patients, 75.53% of chronic patients, and 80.0% of special patients) 76.0% of patients (78.0% of general patients, 80.8% of chronic patients, and 33.3% of special patients) used computer/notebooks or laptops, 48.0% of patients (52.8% of general patients, 47.9% of chronic patients, and 20.0% of special patients) used DVD/CVD and 17.0% of patients (14.3% of general patients, 19.2% of chronic patients, and 20.0% of special patients) used the Mobile Education Kit (MEK) eBooks, web-links, blogs/technical papers and mobile apps. MEK was made to develop a practical teaching and learning environment.\n\nRegarding satisfaction with services at the center, about 53.0% of all patients (66.5% of general patients, 55.6% of chronic patients, and 46.1% of special patients) were highly satisfied with the services; and 42.0% of all patients (40.7% of general patients, 42.6% of chronic patients, and 46.7% of special patients) were extremely satisfied. After completing education program at the center, about 100.0% (100.0% of general patients, 100% of chronic patients, and 100% of disabled patients) continue further education. About 20.0% (22.0% of general patients, 20.2% of chronic patients, and 6.7% of disabled patients) said that the center encouraged them to continue further education. About 48.5% (50.5% of general patients, 49.0% of chronic patients, and 33.3% of disabled patients) rated their satisfaction towards the learning center at a high level; and about 30.0% (27.5% of general patients, 28.7% of chronic patients, and 53.3% of special patients) rated at very high level.\n\n\nDiscussion\n\nFrom the study of the operation of the learning center in the hospital at QSNICH, most child patients have chronic diseases because QSNICH is the only public children’s hospital that serves child patients at the tertiary level and provides care to pediatric patients referred from across Thailand. The most common chronic diseases found in pediatric patients at QSNICH are cancer, heart disease and neurological disease. The results of this study are similar to those of previous studies10,15,16 at tertiary hospitals. Common illnesses in children are gastrointestinal disease, infections, and respiratory illness. Patients with disabilities were children with ADHD, intellectual deficits, learning disabilities and autism. Compared to a previous study10, the number of disabled children at the learning center increased because of the increased medical emphasis on holistic service both physically and developmentally. Teachers at the learning center can follow published guidelines17,18 and utilize a project evaluation manual developed in 2016. The average age of patients with chronic illnesses was 11.22 years, while common illnesses occurred mostly in children with an average age of 9.7 years. followed by an average age of children with disabilities at 9.69 years. After discharge, children in kindergarten were ready to continue their education in primary school.\n\nEducation management should be supported and promoted to encourage students to continue their studies in the regular education system. If a student’s treatment schedule causes them to miss classes, the learning center will modify their study plan by providing extra classes during hospitalization, allowing them to interact with teachers at the learning center. Teachers at the learning center coordinate their curriculum with that of regular school teachers. After being discharged, the children continue their education in the same class. Education19,20, as well as social and behavioral interactions, are very important for children during hospitalization.\n\nThe clinical demographic groups in this study were only selected from children at the Queen Sirikit National Institute of Child Health. The respondents from other institutes should be included as well to increase diversity through a variety of contexts. Future studies should involve learning centers in different types of hospitals such as primary, secondary and tertiary care facilities in each region in Thailand since they all provide different levels of services for patients with different needs.\n\nLearning via social media20–22 are some of the available channels allowing the children to keep in touch with friends and teachers, reduce stress and the feeling of being left out, and prevent social problems after returning to school. They can also share experience at the hospital, turn crisis situations into opportunities and add their own perspectives and ideas about making life better for hospitalized children. Comprehensive19,22 planning will enhance children’s skill in social interaction, education and life adjustment.\n\nOur study found that ten chronically ill children were evaluated, and their non-formal education plans were adjusted appropriately. Similarly, ten disabled children enrolled in IEPs, which provide customized curricula for special needs children that are approved by family and school. Chronically ill and disabled children seem to benefit most from the statistically significant modifications in their education plans. Children with illnesses at the secondary level will be enrolled in non-formal education programs. The study plan of children who stay in the hospital for more than 20 days is more likely to be adjusted than for a five-day period of hospitalization. Other factors studied include the demographics, economic status, occupation and distance from home to school; however, these showed no statistical significance in curriculum adjustment.\n\nFrom the results of our study, it is clear that we should consider evaluating the appropriateness of an education plan by having the teachers at the hospital learning center work closely with regular school teachers to perform long-term monitoring and appropriate planning. The study of parents’ attitudes towards their children’s illnesses and education found that most think chronic diseases are very severe, but parents of special needs children and those with common illnesses do not think this is a serious issue and give more attention to education of their children. Parents of chronically ill children and those with common ailments always inform the school when the patients will be absent, but only about half of the parents of children with disabilities do so because they already have individual education plans and they know that school teachers and administrators will arrange appropriate plans to accommodate the students. Social adjustment19–21,23 is a major concern of parents, most of whom have little confidence that schools can provide adequate services in this area to sick children. On the other hand, parents of disabled children feel moderately to highly secure that the teaching staff at special needs schools will involve their children in building good relationships with friends. In order to arrange learning activities for sick children, teachers must have broad knowledge including educational psychology and a good understanding of the medical conditions because each child has a different set of physical and mental states that require different educational approaches. Most children worry about their illness and this can strongly affect their lifestyle18,21,24 so appropriate guidance is also needed in addition regular coursework.\n\nMobile learning center are good for child patients because their environments are flexible, and they are less likely to feel trapped in a hospital room. Children who must stay in bed will be offered bedside teaching programs. The mobility of educational settings has been greatly improved through use of electronic devices to provide the same standards as obtained in more formal schools. After analyzing the satisfaction result at the learning center, the highest score was from 4.21 to 5.00 (mean = 4.28, SD = 0.62). Considering each item, the respondents reported that having up to date teaching methods and innovative learning activities ranked as the most satisfying features while interesting subject matter and knowledge were second highest because of using technology to help with learning like Electronic Distance Learning Television computer programs to meet international standards as stated by the Ministry of Education.\n\nThe third highest score was enjoyment in the educational entertainment activities, while the fourth ranking went to setting up an environment most conducive to learning. (mean = 4.18, SD = 0.79). Having enough modern learning materials came in fifth. Compared to previous studies10,15, our results demonstrated the importance of adjustments in educational programs as needs changed in line with new teacher guidelines, evaluations of the work of the center, greater involvement of learning materials and technology as standard requirements of the course, and ways to reduce problems between teachers and students with regard to using mobile education. However, new materials, equipment and programs still need to be developed. Based on studies in other countries22,24,25, we concluded that the best educational environment can be achieved with a mix of technology and conventional teaching. The unique, multidisciplinary educational system for sick children in a hospital needs to be supported by hospital, family, and community in concert with local laws. Everyone can support continuing education as a holistic system, reducing the burden on families and society and meeting the purpose of building valuable infrastructure to develop the country’s human resources into the future.\n\n\nConclusions\n\nAll the child patients at the learning center, QSNICH, are able to continue their education. On average, 20% of child patients need to modify their education plan according to their condition. This is most evident among children with chronic illness and disabilities, those at the primary and secondary school levels, and children requiring prolonged hospitalization. The majority of parents have appropriate attitudes about the treatment and education of their children and the children gave the programs their highest rating showing extremely high satisfaction with the service at the learning center in the hospital.\n\n\nData availability\n\nFigshare: Data record of the survey. https://doi.org/10.6084/m9.figshare.8479352.v214.\n\nThis project contains the raw survey responses of each participant.\n\nFigshare: English Survey.pdf. https://doi.org/10.6084/m9.figshare.8208338.v213.\n\nThis project contains the following extended data:\n\n• English Survey.pdf\n\n• QSNICH Princess IT_Version 1.0_09May2017.pdf (survey in Thai)\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nThe authors wish to thank the Committee of the Learning Center in Hospital in the Information Technology Project under the Initiative of Her Royal Highness Princess Maha Chakri Siridhorn and the QSNICH Ethic Committee for support of present study.\n\n\nReferences\n\nStoll BJ, Hansen NI, Bell EF, et al.: Neonatal outcomes of extremely preterm infants from the NICHD Neonatal Research Network. Pediatrics. 2010; 126(3): 443–56. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLocasale Crouch J, Johnson B: Transition from pediatric to adult medical care. Adv Chronic Kidney Dis. 2005; 12(4): 412–7. PubMed Abstract \| Publisher Full Text\n\nVan Cleave J, Gortmaker SL, Perrin JM: Dynamics of obesity and chronic health conditions among children and youth. JAMA. 2010; 303(7): 623–30. PubMed Abstract \| Publisher Full Text\n\nAnnunziato RA, Rakotomihamina V, Rubacka J: Examining the effects of maternal chronic illness on child well-being in single parent families. J Dev Behav Pediatr. 2007; 28(5): 386–91. PubMed Abstract \| Publisher Full Text\n\nWijlaars LP, Gilbert R, Hardelid P: Chronic conditions in children and young people: learning from administrative data. Arch Dis Child. 2016; 101(10): 881–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBell MF, Bayliss DM, Glauert R, et al.: Chronic Illness and Developmental Vulnerability at School Entry. Pediatric. 2016; 137(5): pii: e20152475. PubMed Abstract \| Publisher Full Text\n\nWorld Health Organization: Innovative care for chronic conditions: building blocks for action. Geneva, Switzerland: World Health Organization; 2002 [Internet]. [cited 2018 October15]. Reference Source\n\nCenter for Chronic Disease Prevention and Control: What are chronic and noncommunicable diseases? Ottawa, ON: Public Health Agency of Canada [Internet]. [cited 2018 August 18]. Reference Source\n\nLaw Office of the Basic Education Commission Ministry of Public Health: Education Provision for Persons with Disabilities Act B.E. 2551 and 12 subordinate legislations, Special Education Bureau; 2008. Reference Source\n\nPompimon J: A Study of the organization of instruction programs for young chronically ill in-patient children [Thesis]. Bangkok: Chulalongkorn University; 2000.\n\nEnderlein G: Daniel, Wayne W.: Biostatistics — A Foundations for Analysis in the Health Sciences. Wiley & Sons, New York—Chichester—Brisbane—Toronto—Singapore, 6th ed. 1995, 780 S., £58.—, ISBN 0– 471–58852-0 (cloth). Biom J. 1995; 37(6): 744. Publisher Full Text\n\nYunak R: Interviewing children with chronic illness in a qualitative research. J Nurs Health Sci. 2012; 6: 1–11.\n\nFuengfoo A: English Survey.pdf. figshare. Figure. 2019; http://www.doi.org/10.6084/m9.figshare.8208338.v2\n\nFuengfoo A: Data record of the survey. figshare. Dataset. 2019; http://www.doi.org/10.6084/m9.figshare.8479352.v2\n\nAtheros T, Arpaichiratana C: Model of continuing care for chronically ill children. Thai Journal of Nursing Council. 2011; 26(special issue): 112–25. Reference Source\n\nPugbunmee R: Chronically ill adolescents: impact on self-care. Rama Nurse. 1997; 3: 103–9.\n\nSpecial Education Bureau, Ministry of Education: Practice Guideline for Teachers in Chronic Children Learning Center in Hospital. Bangkok: Office of the Basic Education Commission; 2016.\n\nSpecial Education Bureau, Ministry of Education: Guidelines for Performance Evaluation of Learning Center in Hospital for Chronically Ill Patients. Bangkok: Office of the Basic Education Commission; 2015.\n\nTollit MA, Sawyer SM, Ratnapalan S, et al.: Educational support services for improving school engagement and academic performance of children and adolescents with chronic health condition. Cochrane Public Health. 2015. Reference Source\n\nPerrin EC: Chronic Conditions. In: Parker S, Zuckerman B, editor. Developmental and Behavioral Pediatrics: a handbook for primary care. 2nd edition, Philadelphia: Lippincott Williams & Wilkins; 2005: 152–7.\n\nLum A, Wakefield CE, Donnan B, et al.: Understanding the school experiences of children and adolescents with serious chronic illness: a systematic meta-review. Child Care Health Dev. 2017; 43(5): 645–62. PubMed Abstract \| Publisher Full Text\n\nCouncil for Exceptional Children Division of Physical, Health and Multiple Disabilities: School success for children experience chronic illness: national recommendations for addressing global barriers. Physical Dis. 2017; 36(2): 8–15. Publisher Full Text\n\nQuach J, Barnett T: Impact of chronic illness timing and persistence at school entry on child and parent outcomes: Australian Longitudinal Study. Acad Pediatr. 2015; 15(1): 89–95. PubMed Abstract \| Publisher Full Text\n\nNeto M: Educational motivation meets Maslow: Self-actualization as context driver. Journal of Student Engagement: Education Matters. 2015; 5(1): 18–27. Reference Source\n\nRatnapalan S, Rayar MS, Crawley M: Educational services for hospitalized children. Paediatr Child Health. 2009; 14(7): 433–6. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "57247", "date": "10 Dec 2019", "name": "Issarapa Chunsuwan", "expertise": [ "Reviewer Expertise Developmental and behavioral pediatrics" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe effort to help children with learning center while at the hospital is impressive and the result revealed that many children and families were satisfied with the hospital learning center. I have some suggestions:\nThe title should be adjusted. The word effectiveness should be used to explain what effect the child has when they attended the learning center, such as getting a better score, better adaptation when go back to school. In this research, there is no mention of that issue, but mainly as a study of satisfaction.\n\nAuthors should explain more about the teaching and learning at the center, such as setting and management, teachers: qualification, frequency of attendance, curriculum. Such things would help readers understanding more about the center and could be reproducible.\n\nTable 2: It would be better if authors show or explain more about the concept of what you want to convey? Displaying the percent before and after of each items do not show clear picture. Author may use statistics to analyze or discuss more about Table 2 in the discussion part\n\nTable 3: please check the odd ratio in the illness condition again if the first row (acute illness) is a reference value (odd is equal to 1). The odd value of chronic illness and disability should be greater than 1.\n\nTable 4: The authors try to show the comparison results between 3 groups in degrees of satisfaction. The table should show statistical analysis to compare whether there are differences or not, for clearer picture.\n\nIn discussion part, the author should explain further as specified in the objectives of the research. For example, discuss about satisfaction for learning center, factors effect modification of educational plan.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "61226", "date": "26 Mar 2020", "name": "Savithiri Ratnapalan", "expertise": [ "Reviewer Expertise Education of hospitalized children", "Medical Education", "Paediatric Emergency care" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nGeneral comments The authors sampled children and their parents who had attended a hospital learning center to identify practice patterns, satisfaction and impact. A very good study highlighting the importance of educating hospitalized children and the impact on children and families. The paper needs to be edited again for grammar and spelling.\n\nSpecific Comments\nTable 1 - How often do you explain the lesson to your child? Almost never repeated twice. I believe it should read -Almost always.\nThe discussion part has to be rearranged. After results, the authors can discuss their finding, then either use implications, limitations and conclusions.\nThe first part of the discussion can be moved to the methodology section so that the readers know what common diseases and chronic diseases are being discussed.\nStat discussing the findings first - Your section on implications and recommendations does that but needs to be edited to discuss the study finding.\nThe key messages can be given in a separate text box as bullet points:\n\nEducation of hospitalized children should be supported and promoted to encourage students to continue their studies in the regular education system.\n\nTeachers at hospital learning centers should coordinate their curriculum with regular school teachers and modify the curriculum to suit the child’s educational need.\n\nMobile learning centers can be used for some children to teach at the bed side.\n\nElectronic devices can be used to enhance the education of hospitalized children\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "62258", "date": "12 May 2020", "name": "Banchaun Benjasuwantep", "expertise": [ "Reviewer Expertise Developmental behavioral pediatrician" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe study is about hospital learning center that is interesting and important for children to continue their learning outside school. However, I have some suggestions: There are 200 data in the research rather than 400 people. The author can explain that 200 children included in the research and 200 parents were interviewed.\n\nIntroduction:\n\nThe author should review other countries' hospital learning centers and their effectiveness in education.\n\nMethod:\n\nThe method is short. The author should give more information about the questionnaire especially the relationship between parents and children and children's personal information. The author should explain how it relates to education (the objective of the research) It will make readers clearly understand and follow the article. It is unclear about the hospital learning center such as the reasons teachers modify children's education plans and if they contact children's school teachers?\n\nResult:\n\nThe author should put subtitle before changing the subject, for example, puts words like \"education\" before according to table 1. The readers will clearly follow the article.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1616
https://f1000research.com/articles/8-1615/v1	10 Sep 19	{ "type": "Research Article", "title": "When is early return too early? Clavicle plate fixation in Australian Rules football athletes: A retrospective review of cases", "authors": [ "Matthew Ricks", "Paul Borbas", "Michael Perret", "Sarah Warby", "Greg Hoy", "Paul Borbas", "Michael Perret", "Sarah Warby", "Greg Hoy" ], "abstract": "Background: We retrospectively reviewed return to sport times and results for a series of professional Australian Rules football (AFL) players after clavicle fracture management using a precurved Titanium (Acumed) plate and screws. We allowed very early use and aimed to assess if this construct is strong enough to withstand collision sport activities before radiological union is confirmed. Methods: We reviewed 14 consecutive senior listed AFL players who underwent plate fixation by the senior author over a 10 year period. Outcome measures were taken between 12 and 36 months post operatively and included time to return to play, Nottingham Clavicle Score (NCS) and the Oxford Shoulder Score.\n\nResults: The median time for return to training was 3.5 weeks and 6 weeks for return to full competition. All fractures eventually united. Three of the players that returned before 6 weeks post-operatively suffered significant impacts that caused a bend in the titanium plates. One of these required revision fixation surgically due to perceived weakness in the bent plate construct and cosmetic deformity. The mean Nottingham Clavicle Score and Oxford Shoulder Score were 87.9 and 45.1 points respectively. Survival analysis showed that more than half of the players returned to training at 4 weeks and over 80% at 12 weeks.\n\nDiscussion: The decision to operate early on AFL players who suffered a clavicle fracture in competition play resulted in excellent longterm results. However, the decision to return some players earlier than accepted times for bony union resulted in bending of plates in a significant number, and the risk of further injury must be weighed up in a collision sport. The early return to play time had no adverse effects on performance and participant satisfaction was high, as reflected by the Nottingham Clavicle score and the Oxford shoulder score.", "keywords": [ "Clavicle", "fracture", "fixation", "professional", "sport", "return" ], "content": "Introduction\n\nClavicle fractures are common making up 10% of all fractures and 30% of fractures sustained during sport1,2. Of all sport related fractures, clavicle fractures have the third longest return time to sport3. Clavicle fractures are commonly divided into the anatomical location of fracture (medial, midshaft and lateral) and whether the fracture is undisplaced or displaced. The commonest location of a clavicle fracture is the midshaft in sporting populations4. In a general population, clavicle fractures are managed conservatively or surgically depending on the location, fracture morphology and patient factors5. In athletic individuals, operative intervention is commonly performed for lateral clavicle fractures; however, the management of midshaft fractures is still divided5. Some studies recommend fixation in fractures that are completely displaced, shortened by greater than 2 centimetres or comminution; however, these are based on the non sporting populations and do not transfer across to high level atheletes5.\n\nClavicle fracture fixation carries with it a major complication profile that includes neurovascular injury and infection that can have significant morbidity to the patient. An operative fixation in a displaced fracture of the clavicle has shown improved functional outcomes and a lower malunion and non-union rate compared with nonoperative6–8. Plate fixation has been shown to have a high removal rate electively due to plate irritation9. Studies propose that there is an acceptable margin of safety with fracture fixation when clinically indicated and when performed by a specialist orthopaedic surgeon5,10,11.\n\nAustralian rules football when played professionally in either the Australian Football League (AFL) or the Victorian Football League (VFL) is a physically demanding contact team sport12. Clavicle fractures of the midshaft in an athletic patient are commonly managed surgically, as conservative management causes a reduced ability to return to sport, an increased reinjury rate, scapula dysfunction and poor function secondary to non-union or mal union13. The surgical management of clavicle fractures in contact sports players has been shown to allow safe and early return to sport13.\n\nThere is limited evidence on return to sport time frames post surgically managed clavicle fractures, and no evidence of return to play time frames in AFL and VFL populations. The current return to sport times in surgically managed clavicle fracture in the AFL and VFL are based on expert opinion only. There is a need to provide surgeons, sports physicians, coaches, physiotherapists and players with evidence-based guidelines on safe return to play for AFL and VFL players who have post surgically managed clavicle fractures. A timely and safe return to play post injury is of a high interest to players, clubs and medical practitioners involved in AFL and VFL player care, as a team player lost to injury can affect team dynamics, on field performance and participation satisfaction of the individual13. It is unknown whether a return to a contact sport like AFL with a surgically managed clavicle fracture that has not united results in complications.\n\nThe primary aim of the present study is to retrospectively review the time frames for return to sport, patient reported outcomes measures (PROMS) and complications in a group of professional Australian Rules Football players who were surgically managed for a midshaft clavicle fracture. The secondary aim is to investigate any complications with surgically managed clavicle fractures with an early return to sport before radiographic confirmation of union.\n\n\nMethods\n\nFollowing ethical approval from the institutional review board at Ramsay Health Care (approval: Project 165), a retrospective analysis of all consecutive AFL players who underwent open reduction and internal fixation of a midshaft clavicle fracture was performed. All patients gave written consent for inclusion in the study. The patient data was collected from the Melbourne Orthopaedic Group (Melbourne, Australia) medical database; Ramsay Health Care runs the hospitals that the Lead Surgeon/Author (GH) operates from. We included Australian Rules Football players that played in the professional leagues of the AFL and the VFL who were over the age of 18. Patients who sustained concurrent upper limb injuries (e.g. shoulder dislocation) or who had associated pathology to the clavicle fracture (e.g. ACJ dislocation) were excluded from the study.\n\nAll operations were carried out by the Lead Author (GH) under a general anaesthetic. The surgery is performed in a beach chair position with fluoroscopic assistance. A horizontal skin incision over the midshaft of the clavicle is used and fixation was performed using a precurved Titanium (Acumed) plate. The aim of the surgery is to achieve a rigid fixation construct that would promote union and allow an early return to contact sports. If appropriate a lag screw technique was used and at least 6 cortices engagement with the screws either side of the fracture site.\n\nThe patient left the hospital with a sling and was advised when comfortable to discard and perform early mobilisation of the shoulder. Physiotherapy and exercises were started within a week of discharge with phone consultations and face to face appointments used to promote compliance. The wound had absorbable subcuticular closure with glue sealing the wound. A waterproof dressing was placed over the wound and advised to keep covered till 14 days post-surgery. The patient was advised they were allowed unrestricted movement immediately post op but to refrain from contact sports or weight training till 3 weeks post op.\n\nEach of the patients were followed up and the time taken for each patient to return to sport, any post op complications, or concurrent injuries sustained following the operation were documented. The level the player returned to and whether this was the same league or below was also recorded. The time for return to training and to play along with on field performance scores were collected from the National player database, which is maintained by AFL Australia. A telephone consultation and clinical review was carried out by the Lead Author (GH) collecting the Nottingham clavicle score, Oxford shoulder score and any reported complications from the surgical procedure at six months and one year follow-up.\n\nStatistical analyses were performed using SPSS (v22; IBM). Within-group analyses for the probability of return to training and the probability of return to full competitive play were estimated as a function of time using Kaplan-Meier survival analyses. Survival analyses is used to estimate the time for a particular event to occur within a specified observation time period14. In this this study, the “event” was return to training and then return to full competitive play. Participants who did not have an “event” (did not return to training or competitive play) within the pre-determined observation time period were censored and still accounted for in the analysis to allow for valid inferences14. Censorship means that these participants did not reach their “event” (return to training or return to competitive play) at all or reached their “event” outside the observed time period.\n\nThe observation time period was set to 12 weeks post operation date for return to training and 18 weeks post operation date for return to full competitive play, as we would expect professional AFL players to have returned to training and play within these time periods, even if their operation was performed near or at the end of the season. For return to training, survival tables at 2, 4, 8 and 12 weeks were collated. For return to full competitive play, survival tables at 2, 4, 6 and 18 weeks were collated. Mean and median values with 95% confidence intervals were expressed for time to return to training and time to return to full competition.\n\n\nResults\n\nOf the 98 clavicle fractures that were available on the Melbourne Orthopaedic Group medical database, 15 were of professional AFL players. One was lost to follow up and was excluded from the study leaving 14 patients. In total, 10 of the players injured their left shoulder and 4 injured the right shoulder with 2 of the players injuring their dominant arm. The median age of the players was 28.5 years with a range from 22 to 37 years old (Table 1). The players were operated on between the dates of 2003 to 2017. There were 12 men and 2 women in the study population. A total of 12 of the players were playing in the AFL and 2 were in the VFL and all AFL players returned to their pre injury league of play. The 2 female VFL players did not return to their league level. The mean time to return to training was 4.4 weeks (SD 3.2) with a range of 1.8 to 12. The mean time to return to play was 12.8 weeks (SD 12.8) with a range of 2.5 to 37.5 (Table 2). The median time for return to training was 3.5 weeks and 6 weeks for return to full competition. Figure 1 shows a Kaplan-Meier graph for the return to training and Figure 2 shows the return to same level of play.\n\nNote: * Refers to players who were injured near the end of season and did not return to play until the following season. ** Players who have not yet returned to play.\n\nThe mean Nottingham clavicle score was 87.9 (SD 10.5) with a range of 68 to 100. The mean Oxford shoulder score was 45.1 (SD 4.1) with a range of 35 to 48.\n\nThree patients (21%) sustained a bending of the plate. One of these was on the first game back following the operation and was re-operated on. The other patients had a clinical bending of the plate following a contact but in discussion with the patient they were conservatively managed and went on to unite. One of these patients underwent removal of the plate electively at one year and full fracture union was noted. One patient had a refracture one year post fixation a long time after radiographic union was confirmed. There were no cases of neurovascular injury or infection in this patient group. All of the players returned to training and contact play following the fixation. The complications are displayed in Table 3.\n\n\nDiscussion\n\nThis is the largest series of professional AFL players in the literature showing the return to contact sports following an open reduction and internal fixation of a midshaft clavicle fracture. This patient population is different to the general population due to their high demand and contact nature of their sport. Survival analysis showed that more than half of the players returned to training at 4 weeks and over 80% at 12 weeks. For return to full competitive play, just over half had returned at 6 weeks and this value did not change within the observed time period (18 weeks). One female patient decided to take a season off playing following the injury despite being able to play prior to this and would impact upon the data set. We show in our series that all of the professional AFL players returned to the same level of play post operatively with a mean Oxford shoulder score of 45.1 and a mean Nottingham clavicle score of 87.9. This shows an excellent range of movement and function post operatively.\n\nA study by Verborgt et al. (2005) demonstrated within their semi-professional population (mountain bikers, cyclists, soccer players and swimmers) a 75% patient satisfaction rate who were managed with a fixation of their displaced midshaft clavicle fracture. They showed an infection rate of 18% and a refracture rate of 8% in their professional athletes with 5% nonunion rate. The authors proposed that they showed good results in semi-professional athletes; however, an early return to contact or heavy activities at the expense of significant risk profile would not be considered acceptable in patients with a lower functional demand15. We demonstrated no infections within our study but there were three bent plates (21%) from contacts during play, with one of them requiring a reoperation. This may not have occurred if the population were protected from contact; however, the remaining 11 players (79%) returned to contact before radiographic union and all of the AFL players played at the same level without any structural compromise to the fixation. Despite the three bent plates the players had a very high Oxford shoulder score and Nottingham clavicle score. This study shows that the fixation of the fracture leads to high patient reported outcomes measurements following the surgery but carry the risk of bending of the plates.\n\nA study by Jack et al. (Aug, 2017) showed that surgical fixation of clavicle fractures in professional football players (NFL, USA) lead to decreased union rates, improved shoulder strength and a decreased residual functional impairment. In that series, the mean return to play was 211 days post surgery with 7 (44%) returning in the same season as their injury and subsequent fixation16. Our series of clavicle fixations in collision athletes shows a quicker return to training (4.4 weeks) and playing sport (12.8 weeks). Jack et al. also showed 88.2% of the players remained in the NFL one year after surgery with no difference in post op performance scores or games per season16. In our series, all of the AFL players returned to the same level of play within a year of playing. Another contrasting study by Jack et al (Sep, 2017) showed a return to sport of 96.9% at a mean length of time of 244 days following conservative management of midshaft clavicle fractures in 30 professional NFL players17. That series shows that conservative remains an option in this population; however, it is a slower return when compared with their previous surgical series16 and our study. Robertson and Wood (2016) showed a return rate to sport of 97% with a mean return time of 12 weeks of a midshaft clavicle fracture fixation13, which is a similar length of time to our return to play, but these were noncontact athletes and therefore a different population. Robertson et al. (2018) advocated that surgery in the sports population in a displaced mid shaft clavicle fracture provided an improved return to sport over conservative management. That study did highlight that it was based on low level evidence and that further randomised controlled trials were required to draw more accurate conclusions18. A recent systematic review on the return to sports following clavicle fractures highlighted the difficulty in reporting the return to sports post clavicle fracture due to the lack of detailed studies. The authors of the systematic review advised further studies with a detailed return to sports including time frames, number of weeks missed and if the players returned to their previous level allowing an evidence based recommendation to be13. We have been able to present this data for Australian Football and would encourage other sports to collect the data to allow further evidence and conclusions to be made.\n\n\nConclusion\n\nThe decision to operate early on AFL players who suffered a clavicle fracture in competition play resulted in excellent long term results. However, the decision to return some players earlier than accepted times for bony union resulted in the bending of plates in a significant number of players, and the risk of further injury must be weighed up in a collision sport. The early return to play led to a high Oxford shoulder score and Nottingham clavicle score, resulting in a high level of patient satisfaction. We advocate that plate fixation of mid shaft clavicle fractures is a safe option allowing a very good function and early return to sport, but with the risk of bending of the plate.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nAitken SA, Watson BS, Wood AM, et al.: Sports-related fractures in South East Scotland: an analysis of 990 fractures. J Orthop Surg (Hong Kong). 2014; 22(3): 313–317. PubMed Abstract \| Publisher Full Text\n\nCourt-Brown CM, Wood AM, Aitken S: The epidemiology of acute sports-related fractures in adults. Injury. 2008; 39(12): 1365–72. PubMed Abstract \| Publisher Full Text\n\nRobertson GA, Wood AM, Bakker-Dyos J, et al.: The epidemiology, morbidity, and outcome of soccer-related fractures in a standard population. Am J Sports Med. 2012; 40(8): 1851–7. PubMed Abstract \| Publisher Full Text\n\nAitken S: The Epidemiology of Upper Limb, Lower Limb and Pelvic Fractures in Adults. University of Edinburgh, 2013.\n\nCourt-Brown C, McQueen M, Tornetta P: Trauma. 1st Edition. Orthopaedic Surgery Essentials Series. Philadelphia: Lippincott Williams & Wilkins, 2006. Reference Source\n\nCanadian Orthopaedic Trauma Society: Nonoperative treatment compared with plate fixation of displaced midshaft clavicular fractures. A multicenter, randomized clinical trial. J Bone Joint Surg Am. 2007; 89(1): 1–10. PubMed Abstract \| Publisher Full Text\n\nLädermanna A, Abrassarta S, Denard PJ, et al.: Functional recovery following early mobilization after middle third clavicle osteosynthesis for acute fractures or nonunion: A case-control study. Orthop Traumatol Surg Res. 2017; 103(6): 885–889. PubMed Abstract \| Publisher Full Text\n\nMcKee RC, Whelan DB, Schemitsch EH, et al.: Operative versus nonoperative care of displaced midshaft clavicular fractures: a meta-analysis of randomized clinical trials. J Bone Joint Surg Am. 2012; 94(8): 675–84. PubMed Abstract \| Publisher Full Text\n\nHulsmans MH, van Heijl M, Houwert RM, et al.: High Irritation and Removal Rates After Plate or Nail Fixation in Patients With Displaced Midshaft Clavicle Fractures. Clin Orthop Relat Res. 2017; 475(2): 532–539. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMeisterling SW, Cain EL, Fleisig GS, et al.: Return to athletic activity after plate fixation of displaced midshaft clavicle fractures. Am J Sports Med. 2013; 41(11): 2632–6. PubMed Abstract \| Publisher Full Text\n\nRanalletta M, Rossi LA, Piuzzi NS, et al.: Return to sports after plate fixation of displaced midshaft clavicular fractures in athletes. Am J Sports Med. 2015; 43(3): 565–9. PubMed Abstract \| Publisher Full Text\n\nGabbe B, Finch C: A profile of Australian football injuries presenting to sports medicine clinics. J Sci Med Sport. 2001; 4(4): 386–395. PubMed Abstract \| Publisher Full Text\n\nRobertson GA, Wood AM: Return to sport following clavicle fractures: a systematic review. Br Med Bull. 2016; 119(1): 111–128. PubMed Abstract \| Publisher Full Text\n\nSchober P, Vetter TR: Survival Analysis and Interpretation of Time-to-Event Data: The Tortoise and the Hare. Anaesth Analg. 2018; 127(3): 792–798. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVerborgt O, Pittoors K, Van Glabbeek F, et al.: Plate fixation of middle-third fractures of the clavicle in the semi-professional athlete. Acta Orthop Belg. 2005; 71(1): 17–21. PubMed Abstract\n\nJack RA 2nd, Sochacki KR, Navarro SM, et al.: Performance and Return to Sport After Clavicle Open Reduction and Internal Fixation in National Football League Players. Orthop J Sports Med. 2017; 5(8): 2325967117720677. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJack RA 2nd, Sochacki KR, Navarro SM, et al.: Performance and Return to Sport After Nonoperative Treatment of Clavicle Fractures in National Football League Players. Orthopaedics. 2017; 40(5): e836–e843. PubMed Abstract \| Publisher Full Text\n\nRobertson GA, Wood AM, Oliver CW: Displaced middle-third clavicle fracture management in sport: still a challenge in 2018. Should you call the surgeon to speed return to play? Br J Sports Med. 2018; 52(6): 348–349. PubMed Abstract \| Publisher Full Text" }	[ { "id": "53658", "date": "16 Sep 2019", "name": "Mike Walton", "expertise": [ "Reviewer Expertise Shoulder surgery" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a retrospective cohort review of professional AFL collision athletes undergoing clavicle fracture internal fixation.\nThe authors report very early return to play before the progression to osseous union.\nThe authors include 14 patients. However 2 patients (13 & 14) have not yet returned to play. If these patients are to be included in the series and explanation of the lack of return to either training or play must be included. If this information is not available they should be excluded from analysis.\nThe authors identify that the majority (80%) were able to return to training at 4-6 weeks without complications. However 20% suffered plate damage. This is an avoidable complication such that 1 patient in the series required reoperation. This patient (12) did not return to play for 28 weeks, far longer than would have been necessary. Little detail is provided as to what happened after return to training/play in the groups. I doubt that this information is available but it is highly likely that those patients who had plate failure did so because of greater impacts that those who did not. Returning to training in the early phase is likely to be safe if you can avoid heavy impact on the shoulder until union has occurred. This is likely to be influenced by the position of the player (known) and the frequency of collisions (unknown).\nThe authors have documented that the bent plates were managed conservatively in patients 8 and 14. How long were they kept out of play for after the bending occurred? This is important as it would highlight the implications of the reinjury on end return to play time. From the data patient 14 has not yet returned to play, 12 took 28.5 weeks. Patient 8 returned to play at 6 weeks but we do not know when the plate bent and whether he required a further period of rest after the initial return to play.\nWhilst the authors conclude that this study shows that early return to play is possible, it shows the high rate of significant avoidable harm that this approach entails. These risks may be taken on by the patient after appropriate education but the authors cannot conclude (as in the discussion) that early return to play had no adverse effects on performance. It should be stated clearly that if plate failure does occur then overall return to play is significantly longer.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "53657", "date": "03 Oct 2019", "name": "Benjamin Dean", "expertise": [ "Reviewer Expertise Upper limb surgery and pain" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nOverall:\nThis is a useful piece of work which challenges the way in which clavicle fractures may be rehabilitated following surgery in athletes.\n\nSpecifics:\nAbstract:\nI would re-write the background as some of this relates to methods and not background. Just frame the rationale for this study.\n\nDo you know what the time to union was? (May be useful to put this in the abstract).\n\nMain paper:\nIntroduction:\nMay be worth stating what is roughly standard practice in athletes? If there is evidence here quote it, if not then state what is generally done – I suspect the standard practice is to wait for evidence of some bony union before allowing contact, hence why this study is interesting as it questions current practice.\n\nAims:\nYou mention complications in both primary and secondary aims, can you refine this?\n\nResults:\nWould be worth adding in time to union if you have this?\n\nTable 1 – would be worth adding another row with the median/average values of age etc\n\nDiscussion:\nI would re-do the first paragraph to concisely summarise your key results and why this is important, i.e. that early return to full contact is possible and there is a risk/benefit to this approach as shown by your results.\n\nConclusion:\nI would make this a bit more concise to summarise that early return to full contact is feasible and results in excellent outcomes but there is the risk of…….\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1615
https://f1000research.com/articles/7-534/v1	02 May 18	{ "type": "Research Article", "title": "Validation of EuroSCORE II in patients undergoing coronary artery bypass grafting (CABG) surgery at the National Heart Institute, Kuala Lumpur: a retrospective review.", "authors": [ "Ahmad Farouk Musa", "Xian Pei Cheong", "Jeswant Dillon", "Rusli Bin Nordin", "Xian Pei Cheong", "Jeswant Dillon", "Rusli Bin Nordin" ], "abstract": "Background: The European System for Cardiac Operative Risk (EuroSCORE) II was developed in 2011 to replace the aging EUROScore for predicting in-house mortality after cardiac surgery. Our aim was to validate EuroSCORE II in Malaysian patients undergoing coronary artery bypass graft (CABG) surgery at our Institute. Methods: A retrospective single-center study was performed. A database was created to include EuroSCORE II values and actual mortality of 1718 patients undergoing CABG surgery in Malaysia from 1st January to 31st December 2016. The goodness-of-fit of EuroSCORE II was determined by the Hosmer-Lemeshow goodness-of-fit test and discriminatory power with the areas under the receiver operating characteristics (ROC) curve (AUC). Results: Observed mortality rate was 4.66% (80 out of 1718 patients). The median EuroSCORE II value was 2.06% (Inter Quartile Range: 1.94%) (1st quartile: 1.45%, 3rd quartile: 3.39%). The AUC for EuroSCORE II was 0.7 (95% CI 0.640 – 0.759) indicating good discriminatory power. The Hosmer-Lemeshow goodness-of-fit test did not show significant difference between expected and observed mortality in accordance to the EuroSCORE II model (Chi-square = 13.758, p = 0.089) suggesting good calibration of the model in this population. Cross-tabulation analysis showed that there is slight overestimation of EuroSCORE II in low-risk groups (0-10%) and slight underestimation in high-risk groups (>20%). Multivariate logistic regression analysis showed that gender, age, total hospital stay, serum creatinine and critical pre-operative state are significant predictors of mortality post-CABG surgery. Conclusion: This study indicated that the EuroSCORE II is a good predictor of post-operative mortality in the context of Malaysian patients undergoing CABG surgery. Our study also showed that certain independent variables might possess higher weightage in predicting mortality among this patient group. Therefore, it is suggested that EuroSCORE II can be safely used for risk assessment while ideally, clinical consideration should be applied on an individual basis.", "keywords": [ "EuroSCORE II", "predictor", "coronary artery bypass graft", "post-CABG mortality", "National Heart Institute" ], "content": "Introduction\n\nCoronary Artery Bypass Grafting (CABG) surgery, being a major surgery, is not without significant risks, up to and including death. In the United States, operative death rate and in-hospital mortality rate post CABG between 1997 and 2001 ranged from 1% to 5% for all patients1,2. In Malaysia, statistics from the National Heart Institute (IJN) had shown that the mortality rate for patients undergoing CABG surgery in Malaysia was around 2.7%3. Notwithstanding, it is important to take note that the associated risk is very much dependent on multiple interacting factors including patients’ comorbidities and occurrence of any complications due to the operation itself4,5.\n\nThe need for a simple tool to predict post-surgical mortality led to the development of the European System for Cardiac Operative Risk Evaluation (EuroSCORE), also known as the European System for Cardiac Operation Risk Evaluation in 1999. This is a risk evaluation tool to calculate and predict operative mortality in patients undergoing cardiac surgery. It was developed using risk factors collected from almost 20,000 patients from more than 100 hospitals in Europe6.\n\nSince the publication of EuroSCORE, it had been widely employed and validated in various populations of cardiac surgical patients. However, it was found that the additive score for EuroSCORE tended to underestimate the risk of mortality, possibly when there were co-existing risk factors in high-risk patients. These concerns led to the development of the more complicated logistic EuroSCORE I. This version did produce a better estimate of risk in high risk patients. However, its main drawback is the overestimation of risk despite improvements in cardiac surgical outcomes observed7.\n\nIn order to overcome this issue, the EuroSCORE team has come up with a revised version, which is known as the EuroSCORE II during the 2011 EACTS meeting in Lisbon. EuroSCORE II was developed by collecting and analysing prospective risk and outcome data on 22,381 patients undergoing major cardiac surgery in 154 hospitals in 43 countries over a 12-week period (May-July 2010). The new EuroSCORE II has updated the definition of renal function and unstable angina. Also, it further subdivided the classification of pulmonary hypertension, urgency and weight of operation. Most importantly, the new model has also changed the definition of outcome measurement, from 30-day mortality rate to in-hospital mortality. The main reason was the loss of follow-up data after discharge in certain centres, thereby giving rise to poor quality data in the original EuroSCORE8.\n\nThroughout the years, multiple validation studies have been conducted around the world including Europe, America and Asia to examine the validity of EuroSCORE II in predicting post-operative mortality and it had shown different results regarding the discriminatory power and calibration of this scoring system in different populations.\n\nFurthermore, the EuroSCORE II has yet to be validated in Malaysia, a country with high incidence of cardiovascular diseases. Therefore, this study will serve as the first in Malaysia to examine the validity of EuroSCORE II in predicting operative mortality among patients undergoing CABG surgery.\n\n\nMethods\n\nA single-centre retrospective review was conducted at the National Heart Institute (IJN), the largest heart center in Malaysia. Almost all of the information needed was retrieved from the IJN electronic in-house database. Out-of-hospital data including death and late complications was obtained via telephone enquiry.\n\nEthical approval was obtained from both the IJN Research Ethics Committee (IJNREC/238/2018) and Monash University Human Research Ethics Committee (MUHREC/12981). The study was also registered with the National Medical Research Register (NMRR-17-2749-39322).\n\nWithin the period from 1st January 2016 to 31st December 2016, 1718 consecutive patients undergoing CABG surgery at the IJN were included in this study.\n\nInclusion Criteria: All patients undergoing CABG including two or three procedures.\n\nExclusion Criteria: Reinterventions for any cause in the same admission as the primary operation.\n\nEuroSCORE II included ten patient-related factors, five cardiac-related factors, and three operation-related factors. Patient related factors include age (year), gender (male /female), renal impairment (creatinine clearance), extracardiac arteriopathy, poor mobility, previous cardiac surgery, chronic lung disease, active endocarditis, critical preoperative state and diabetes on insulin. Cardiac related factors include the New York Heart Association (NYHA) stages, Canadian Cardiovascular Society (CCS) class 4 angina, Left Ventricular (LV) function (ejection fraction > 50%, 31-50%, 21-30%, <20%), recent myocardial infarction (MI) (within 90 days) and pulmonary hypertension (31-55 mm Hg / >55 mm Hg). Operation related factors include urgency (elective, urgent, emergency, salvage), weight of the intervention (isolated CABG, isolated single non-CABG, 2-procedures, 3-procedures) and surgery on thoracic aorta. Details regarding EuroSCORE II calculation are available from the EuroSCORE site. The outcome variable, which is in-hospital mortality, was retrieved from the in-hospital database. In other words, it simply means death occurring at any time after surgery during the current admission. Additionally, important clinical information including presence of comorbidities (hypertension and hypercholesterolemia), total hospital stay, total ICU stay and follow-up status were also collected. A database is then created to collect the relevant data and stored in spreadsheets.\n\nData was evaluated using the Microsoft Excel 2016 database (Microsoft Inc.) and analyzed using the Statistical Package for Social Sciences (SPSS) version 23.0. Continuous variables were presented as mean and standard deviation. Categorical variables were presented as frequencies and compared between groups using the chi-square test. A multiple logistic regression analysis was undertaken to determine significant predictors of in-hospital mortality. Predictive ability of the estimation model was assessed through discriminatory power and calibration. Receiver operating characteristics (ROC) curve analysis was performed to estimate the discriminant ability of this risk scoring model in predicting immediate post-operative mortality. It was considered good if the area under the curve (AUC) was >0.70. Calibration was evaluated using the Hosmer-Lemeshow goodness-of-fit test.\n\n\nResults\n\nThe demographics and pre-operative characteristics of patients are shown in Table 1. In terms of social demographics, mean age was 60 ± 8.89 years old, women made up 15.9% of the total sample, and Malay constituted the largest ethnic group (53.8%), which corresponds to the race distribution in Malaysia. Majority of the patients had comorbidities such as hypertension (83.3%) and hypercholesterolemia (77.4%). Preoperatively, the majority of patients were in NYHA class I (41.2%) and II (49.9%). Majority (46.6%) had good left ventricular function. Intraoperatively, majority of patients underwent isolated CABG (86.6%) without previous history of cardiac surgery (98.7%).\n\nn: Number of patients, SD: Standard deviation, CCS: Canadian Cardiovascular Society, MI: Myocardial infarction, NYHA: New York Heart Association, EF: Ejection fraction, CABG: Coronary artery bypass grafting, PA: Pulmonary artery, EuroSCORE: European system for cardiac operative risk evaluation\n\nThe actual in-hospital mortality rate was 4.7% (80 out of 1718 patients). In comparison, predicted mortality rate by the median EuroSCORE II value was 2.06 (1st quartile: 1.452, 3rd quartile: 3.389). In other words, the predicted in-hospital mortality rate was slightly lower compared to the observed mortality rate. The correct classification was seen for 1638 out of 1718 patients, giving rise to a success rate of 95.3%. Actual mortality rate, by quartiles of EuroSCORE II, was 1.6% in the first quartile, 3.0% in the second quartile, 4.7% in the third quartile and 9.4% in the fourth quartile as shown in Table 2.\n\nEuroSCORE: European system for cardiac operative risk evaluation\n\nAs illustrated in Figure 1, the area under the receiver operating characteristic curve (AUC) was 0.7 (95% CI 0.640 – 0.759, p < 0.001), suggesting that the EuroSCORE II has fair and acceptable discriminatory power to discriminate between incidences of patients who died and those who were alive.\n\nThe Hosmer-Lemeshow (HL) goodness-of-fit test did not show significant difference between expected and observed mortality in accordance to the EuroSCORE II model (Chi-square: 13.748, p = 0.089), indicating reasonable calibration of this model in predicting in-hospital mortality among patients who underwent CABG surgery. Cross-tabulation analysis of predicted risk by EuroSCORE II showed that there was slight overestimation in low risk group (0 – 10%) and slight underestimation in high risk group (>20%) as shown in Table 3. Figure 2 shows the relationship between age and EuroSCORE II in patients post-CABG in the IJN, Malaysia.\n\nEuroSCORE: European system for cardiac operative risk evaluation\n\nAnalysis on discriminatory power were repeated on various subgroups, including gender, race (Malay, Chinese, Indian and others), age (below 60 years old, 60 years old and above) and presence of comorbidities (hypertension, diabetes mellitus and hypercholesterolemia). Subgroup analysis showed that AUC for the EuroSCORE II was 0.695 in male (95% CI 0.620 – 0.770, p < 0.001), 0.642 in female (95% CI 0.534 – 0.751, p = 0.017), 0.696 in Malay (95% CI 0.624 – 0.767, p < 0.001), 0.801 in Chinese (95% CI 0.696 – 0.906, p < 0.001), 0.642 in Indians (95% CI 0.470 – 0.813, p = 0.083), 0.596 in other ethnicities (95% CI 0.153 – 1.000, p = 0.573), 0.700 in those aged below 60 years old (95% CI 0.571 – 0.830, p = 0.006), 0.673 in those age 60 years old and above (95% CI 0.603 – 0.743, p < 0.001), 0.691 in hypertensive patients (95% CI 0.628 – 0.754, p < 0.001), 0.745 in patients without hypertension (95% CI 0.585 – 0.905, p = 0.04), 0.672 in diabetic patients (95% CI 0.602 – 0.741, p < 0.001), 0.728 in patients without diabetes (95% CI 0.614 – 0.841, p < 0.001), 0.683 in patients with hypercholesterolemia (95% CI 0.615 – 0.750, p < 0.001) and 0.768 in patients without hypercholesterolemia (95% CI 0.650 – 0.886, p < 0.001).\n\nMultivariate binary logistic regression analysis was undertaken to develop a prediction model of variables in EuroSCORE II and outcome (in-hospital mortality). The forward conditional method was selected to be used for analysis. The last step showed that being female, aged more than or equal to 65 years old, serum creatinine more than 120 micromole/litre and longer ICU stay are significant and independent predictors of in-hospital mortality in patients undergoing CABG surgery as shown in Table 4. The model which consists of the four risk factors explained between 24.6% (Cox and Snell R-square) and 47.8% (Nagelkerke R-square) of variance in predicting in-hospital mortality among patients undergoing CABG surgery in Malaysia. Moreover, it correctly classified 99% of the cases. These independent variables also made a unique statistically significant contribution to the model (χ2 (8) = 92.403, p < 0.001).\n\n\nDiscussion\n\nAccurate prediction of risk is always essential and plays an important role in guiding doctors to make clinical decision as to whether surgery is an appropriate intervention, especially among high risk patients. In the field of cardiothoracic surgical practice, several risk assessment tools or models, including the EuroSCORE II, have been proposed and developed by researchers based on clinical databases selected from specific populations8. Concurrently, the EuroSORE II has become one of the most commonly used risk evaluation tool in many cardiac centres worldwide. However, it is crucial to note that the EuroSCORE II was actually developed based on data from mainly European countries9. Therefore, the application of EuroSCORE II in other populations might need cautious clinical consideration as there are other inter-related factors such as genetic background of the population, different healthcare systems as well as different social and cultural practice.\n\nIn our present study, we have determined both the calibration and discriminatory power of the EuroSCORE II in our local population undergoing CABG surgery. Calibration of a model includes the determination of its ability to compare the predicted outcome (EuroSCORE II) with the actual outcome (actual in-hospital mortality) in the entire sample. Discriminatory power is the ability of the EuroSCORE II to distinguish patients who were still alive or who died in the hospital. Results showed that the EuroSCORE II has reasonable and fair calibration and discriminatory power in this group of Malaysian patients who underwent CABG surgery.\n\nMany studies have been conducted around the world to examine the validity of the EuroSCORE II in predicting in-hospital mortality post-CABG. First of all, one of the most important findings in our study will be the in-hospital mortality rate. According to multiple validated studies that had been conducted across Europe, it showed a mortality rate ranging from 3%, 3.7%, and 4.85% in Italy10, Greece11 and Serbia12, respectively. Consistent with previous literature, we observed an actual in-hospital mortality rate of 4.7% in this study.\n\nA multicentre prospective validation study was done to compare the EuroSCORE and EuroSCORE II among 4000 patients undergoing cardiac surgery in Spain. Results showed that both the EuroSCORE and EuroSCORE II has good discriminatory power (AUC > 0.75). In addition, the original EuroSCORE tends to over-predict mortality while EuroSCORE II under-predict mortality13. Similarly, a single centre validation study in Hungary has also shown that EuroSCORE overestimated the risk of mortality while EuroSCORE II underestimated the risk. Despite that, EuroSCORE II was still better than its original version in terms of discriminatory power14. Our study results showed a skewed distribution of EuroSCORE II and the median was 2.06 in comparison to the actual mortality rate of 4.7%, which certainly showed underestimation of risk by the EuroSCORE II.\n\nIn terms of calibration and discriminatory power, most of the validation studies in Europe including Spain, Italy, Greece, Serbia and Hungary has an AUC of more than 0.7, which indicates good discriminatory power and calibration10–14. However, there was a collaborative study between two centres in the Netherlands and United Kingdom, which showed that EuroSCORE II was not good in predicting mortality in patients undergoing cardiac operation. It showed an unsatisfactory AUC of 0.67, indicating poor discriminatory power. Particularly in middle-eastern countries, a slightly different results were observed. For instance, in Pakistan, it was shown that, despite having a satisfactory discriminatory power, EuroSCORE II was poorly calibrated and the original EuroSCORE actually fared better than the EuroSCORE II among isolated CABG patients in their local population15. This can be attributed to various demographic-related factors or even study bias. Among our population of CABG patients, we observed an AUC of 0.7, which is deemed to be satisfactory in predicting in-hospital mortality.\n\nOur study had shown that only female gender, age more than or equal to 65 years old, serum creatinine more than 120 micromole/litre and longer ICU stay are significant predictors of in-hospital mortality in patients post CABG surgery. In this context, independent variables were selected in line with the principle of parsimony so that our analysis can be more consistent and limited to as few variables as possible in the prediction model.\n\nAccording to previous literatures, increasing age was found to be a significant risk factor by a few studies to investigate age as a risk predictor in patients undergoing CABG surgery16,17. In terms of gender, multiple studies had shown that female gender was an independent predictor for early and late mortality after cardiac operation18–20. Chronic renal dysfunction has also been known to have close association with mortality after CABG. After the establishment of EuroSCORE in 2003, a study was performed to look into patients undergoing CABG with a preoperative serum creatinine <200 µmol/L. It was shown that both the in-hospital mortality rate and stroke rate for this group of patients went up to 2.5%. Furthermore, the mortality rate also increased with increasing preoperative serum creatinine level21.\n\nRisk prediction is a very important area in cardiothoracic surgery that can serve to further refine the quality of patient care. By taking into consideration a series of relevant risk factors, the predicted risk by EuroSCORE II can guide us as to whether to perform an operation or to treat conservatively certain patients. Given the fact that multiple studies had shown that the original EuroSCORE was outdated and not applicable for risk prediction7,22, EuroSCORE II can replace its predecessor as a risk prediction model for mortality prediction. As discussed previously, a significant number of cardiac centres around the world including Europe, Asia and the Middle-East had validated EuroSCORE II with acceptable results. We believe that it can serve as a practical tool for the benefits of cardiac surgeons in terms of risk analysis, quality assurance as well as cost consideration.\n\nNonetheless, we do not deny the fact that it is still virtually impossible to develop an ideal risk evaluation model that fits everyone in the world as all of the models were developed based on clinical data from certain region-specific population. Moreover, given that cardiac surgery has gone through major advancement over the years in terms of improvement in surgical techniques and perioperative care, preoperative risk prediction has been shown to be a moving target that is both important and challenging to tackle.\n\nLooking forward, our efforts for improvement will focus on the universality and practicability of the risk evaluation model. First of all, the lack of parsimony is a problem with the EuroSCORE II, which consists of 18 variables. A simpler risk prediction model with fewer variables that is able to predict in-hospital mortality would be better23,24. Should we be able to develop a relatively simpler and straightforward risk model in the future, the aim will be to have it provide the same predictive power but also be more user-friendly. Following that, we also recommend that a multicentre large scale study should be undertaken to incorporate population groups from all over the world with more variation in terms of genetic and social backgrounds so that a universal and culturally sensitive risk assessment model can be developed in the future.\n\n\nLimitations\n\nThis study was limited by its nature of retrospective study. There was a considerable amount of missing data, which might lead to a relatively smaller sample size in performing logistic regression analysis on various independent risk factors. Due to its retrospectivity, patients with specific risk groups cannot be intentionally selected. In our case, we observed a skewed distribution of patients in terms of risk group (more than 90% of our patients are within the low risk group of 0-10%). In addition, Peterson et al.4 from the Duke Clinical Research Institute had looked into the association between surgeon experience and mortality post-CABG. It was shown that surgeon experience was a significant predictor of mortality. The highest mortality rate was observed when patients were treated by low-volume surgeons. This study was conducted in a cardiac centre with surgeons with varying levels of surgical experience. That might directly or indirectly affect the outcome of surgery or even in-hospital mortality to a certain extent.\n\n\nConclusion\n\nThis single centre large validation study showed that the EuroSCORE II exhibits reasonable and fair discriminatory power and calibration in predicting in-hospital mortality risk among patients undergoing CABG surgery in Malaysia. Despite being a single centre study and therefore may not be representative of the entire population, we think that it can be safely used as a risk assessment tool with cautious clinical consideration being applied on an individual basis.\n\n\nData availability\n\nRaw data for the study ‘Validation of EuroSCORE II in patients undergoing coronary artery bypass grafting (CABG) surgery in Malaysia’ are available both in excel and SAV formats. Data analysis is available in SPV format (SPSS output).\n\nDataset 1 – Demography and EUROScore II variables (Excel) 10.5256/f1000research.14760.d20221525\n\nDataset 2 – Demography and EUROScore II variables (SPSS) 10.5256/f1000research.14760.d20221626\n\nDataset 3 – SPSS Output 10.5256/f1000research.14760.d20221727\n\n\nEthics approval\n\nAll research procedures were done in accordance with the ethical regulations set by the IJN ethics committee, Monash University Human Research Ethics Committee (MUHREC) and it abides with the Helsinki Declaration revised in 2013.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgement\n\nThe authors would like to extend our gratitude to all the staff at the Research Department, National Heart Institute, especially Ms. Norfazlina Binti Jaffar, assistant manager of the data management & biostatistical support services, for their assistance in data collection.\n\n\nReferences\n\nHannan EL, Racz MJ, Walford G, et al.: Long-term outcomes of coronary-artery bypass grafting versus stent implantation. N Engl J Med. 2005; 352(21): 2174–2183. PubMed Abstract \| Publisher Full Text\n\nBirkmeyer JD, Siewers AE, Finlayson EV, et al.: Hospital volume and surgical mortality in the United States. N Engl J Med. 2002; 346(15): 1128–37. PubMed Abstract \| Publisher Full Text\n\nMusa AF, Chao ZQ, Low ZX, et al.: Retrospective review on atrial fibrillation after coronary artery bypass grafting: a single centre experience. Paper presented at: 40th World Congress of the International College of Surgeons; 2016; Oct 23–26; Kyoto, JP.\n\nPeterson ED, Coombs LP, Delong ER, et al.: Procedural volume as a marker of quality for CABG surgery. JAMA. 2004; 291(2): 195–201. PubMed Abstract \| Publisher Full Text\n\nCram P, Rosenthal GE, Vaughan-Sarrazin MS: Cardiac revascularization in specialty and general hospitals. N Engl J Med. 2005; 352(14): 1454–1462. PubMed Abstract \| Publisher Full Text\n\nRoques F, Nashef SA, Michel P, et al.: Risk factors and outcome in European cardiac surgery: analysis of the EuroSCORE multinational database of 19030 patients. Eur J Cardiothorac Surg. 1999; 15(6): 816–22; discussion 822–3. PubMed Abstract \| Publisher Full Text\n\nNashef SA; EuroSCORE Project Team: The New EuroSCORE Project. Nowa skala EuroSCORE. Kardiol Pol. 2010; 68(1): 128–129. PubMed Abstract\n\nNashef SA, Roques F, Sharples LD, et al.: EuroSCORE II. Eur J Cardiothorac Surg. 2012; 41(4): 734–44; discussion 744–5. PubMed Abstract \| Publisher Full Text\n\nGranton J, Cheng D: Risk stratification models for cardiac surgery. Semin Cardiothorac Vasc Anesth. 2008; 12(3): 167–174. PubMed Abstract \| Publisher Full Text\n\nPaparella D, Guida P, Di Eusanio G, et al.: Risk stratification for in-hospital mortality after cardiac surgery: external validation of EuroSCORE II in a prospective regional registry. Eur J Cardiothorac Surg. 2014; 46(5): 840–848. PubMed Abstract \| Publisher Full Text\n\nStavridis G, Panaretos D, Kadda O, et al.: Validation of the EuroSCORE II in a Greek Cardiac Surgical Population: A Prospective Study. Open Cardiovasc Med J. 2017; 11: 94–101. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNezic D, Spasic T, Micovic S, et al.: Consecutive Observational Study to Validate EuroSCORE II Performances on a Single-Center, Contemporary Cardiac Surgical Cohort. J Cardiothorac Vasc Anesth. 2016; 30(2): 345–351. PubMed Abstract \| Publisher Full Text\n\nGarcia-Valentin A, Mestres CA, Bernabeu E, et al.: Validation and quality measurements for EuroSCORE and EuroSCORE II in the Spanish cardiac surgical population: a prospective, multicentre study. Eur J Cardiothorac Surg. 2016; 49(2): 399–405. PubMed Abstract \| Publisher Full Text\n\nKoszta G, Sira G, Szatmári K, et al.: Performance of EuroSCORE II in Hungary: a single-centre validation study. Heart Lung Circ. 2014; 23(11): 1041–1050. PubMed Abstract \| Publisher Full Text\n\nQadir I, Alamzaib SM, Ahmad M, et al.: EuroSCORE vs. EuroSCORE II vs. Society of Thoracic Surgeons risk algorithm. Asian Cardiovasc Thorac Ann. 2014; 22(2): 165–171. PubMed Abstract \| Publisher Full Text\n\nRocha AS, Pittela FJ, Lorenzo AR, et al.: Age influences outcomes in 70-year or older patients undergoing isolated coronary artery bypass graft surgery. Rev Bras Cir Cardiovasc. 2012; 27(1): 45–51. PubMed Abstract \| Publisher Full Text\n\nNaughton C, Feneck RO, Roxburgh J: Early and late predictors of mortality following on-pump coronary artery bypass graft surgery in the elderly as compared to a younger population. Eur J Cardiothorac Surg. 2009; 36(4): 621–7. PubMed Abstract \| Publisher Full Text\n\nBlankstein R, Ward RP, Arnsdorf M, et al.: Female gender is an independent predictor of operative mortality after coronary artery bypass graft surgery: contemporary analysis of 31 Midwestern hospitals. Circulation. 2005; 112(9 Suppl): I323–7. PubMed Abstract\n\nBarbir M, Lazem F, Ilsley C, et al.: Coronary artery surgery in women compared with men: analysis of coronary risk factors and in-hospital mortality in a single centre. Br Heart J. 1994; 71(5): 408–12. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEdwards FH, Carey JS, Grover FL, et al.: Impact of gender on coronary bypass operative mortality. Ann Thorac Surg. 1998; 66(1): 125–31. PubMed Abstract \| Publisher Full Text\n\nZakeri R, Freemantle N, Barnett V, et al.: Relation between mild renal dysfunction and outcomes after coronary artery bypass grafting. Circulation. 2005; 112(9 Suppl): 1270–1275. PubMed Abstract\n\nZingone B, Pappalardo A, Dreas L: Logistic versus additive EuroSCORE. A comparative assessment of the two models in an independent population sample. Eur J Cardiothorac Surg. 2004; 26(6): 1134–1140. PubMed Abstract \| Publisher Full Text\n\nWells CK, Feinstein AR, Walter SD: A comparison of multivariable mathematical methods for predicting survival--III. Accuracy of predictions in generating and challenge sets. J Clin Epidemiol. 1990; 43(4): 361–372. PubMed Abstract \| Publisher Full Text\n\nRanucci M, Castelvecchio S, Menicanti L, et al.: Risk of assessing mortality risk in elective cardiac operations: age, creatinine, ejection fraction, and the law of parsimony. Circulation. 2009; 119(24): 3053–3061. PubMed Abstract \| Publisher Full Text\n\nMusa AF, Cheong XP, Dillon J, et al.: Dataset 1 in: Validation of EuroSCORE II in patients undergoing coronary artery bypass grafting (CABG) surgery at the National Heart Institute, Kuala Lumpur: a retrospective review. F1000Research. 2018. Data Source\n\nMusa AF, Cheong XP, Dillon J, et al.: Dataset 2 in: Validation of EuroSCORE II in patients undergoing coronary artery bypass grafting (CABG) surgery at the National Heart Institute, Kuala Lumpur: a retrospective review. F1000Research. 2018. Data Source\n\nMusa AF, Cheong XP, Dillon J, et al.: Dataset 3 in: Validation of EuroSCORE II in patients undergoing coronary artery bypass grafting (CABG) surgery at the National Heart Institute, Kuala Lumpur: a retrospective review. F1000Research. 2018. Data Source" }	[ { "id": "33778", "date": "13 Jun 2018", "name": "Hasanat Sharif", "expertise": [ "Reviewer Expertise Clinical outcome research" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors have used same methodology as we did in our study. This is the usual standard.\nPlease share details of two and three procedures whether they were CABG AVR or CABG MVR or DVR!\n\nIt would be a better idea to analyse mortality in performance of Euroscore II for isolated CABG and combined procedure separately.\n\nTable 3 there is concordance between expected and observed mortalities in subgroups 0-10% and 11-20%. However for >20% there is gross mismatch of expected and observed mortalities. Please share details of 8 patients who died in this group. Did they belong to isolated CABG category or the combined procedure? As suggested above, please separate analysis of performance of Euroscore II for isolated CABG and combined procedures. This will give a true picture of performance in these two distinct groups. It may also explain the reason for underestimation in >20% group and may even improve concordance within results.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "3755", "date": "22 Jun 2018", "name": "Ahmad Farouk Musa", "role": "Author Response", "response": "Dear Prof Hasanat, thank you for your comments. My answers were detailed below accordingly. Please share details of two and three procedures whether they were CABG AVR or CABG MVR or DVR In our study, there are 160 patients who underwent a combination of two procedures, which included CABG + AVR, CABG + MVR as well as CABG + Aortic Root Replacement. For 33 patients who underwent three procedures, it included CABG+MVR+Aortic Root Replacement; CABG+MVR+AVR; as well as CABG + ASD Closure + Devega’s Tricuspid Annuloplasty. It would be a better idea to analyze mortality in the performance of Euroscore II for isolated CABG and combined procedure separately. Thank you again for your suggestion. We have already performed subgroup analysis based on weightage of procedures. Among 1488 patients who underwent isolated CABG, we observed an actual in-hospital mortality rate of 3.9% and a median EuroSCORE II (predicted mortality) of 1.918, which showed an underestimation of risk. Hosmer and Lemeshow Goodness-of-Fit test showed a significant p value of 0.032, indicating a significant difference between expected and actual mortality among this group of patients. Discriminatory power as shown by the ROC curve analysis showed an area of 67.8%. Among 160 patients who underwent two procedures, we observed an actual in-hospital mortality rate of 10% and a median EuroSCORE II (predicted mortality) of 3.712, which showed an underestimation of risk. However, Hosmer and Lemeshow Goodness-of-Fit test showed a non-significant p-value of 0.591, indicating no significant difference between expected and actual mortality among this group of patients. Discriminatory power as shown by the ROC curve analysis showed an area of 63.4%. Lastly, among 33 patients who underwent three procedures, we observed an actual in-hospital mortality rate of 18.2% and a median EuroSCORE II (predicted mortality) of 6.2, which showed an underestimation of risk. Similar to two procedures, Hosmer and Lemeshow Goodness-of-Fit test showed a non-significant p-value of 0.575, indicating no significant difference between expected and actual mortality among this group of patients. Discriminatory power as shown by the ROC curve analysis showed an area of 67.9%. Table 3 there is concordance between expected and observed mortalities in subgroups 0-10% and 11-20%. However, for >20% there is a gross mismatch between expected and observed mortalities. Please share details of 8 patients who died in this group. Did they belong to isolated CABG category or the combined procedure? Among the 8 patients who died in the 11-20% subgroup, 4 of them had isolated CABG (50%), while 2 of them underwent two procedures (25%) and the other 2, three procedures (25%) respectively." } ] }, { "id": "48306", "date": "13 May 2019", "name": "Domingo Marcolino Braile", "expertise": [ "Reviewer Expertise Cardiovascular Surgery", "Emeritus Professor at FAMERP and UNICAMP. Editor in Chief of Brazilian Journal OF Cardiovascular Surgery [www.bjcvs.org]" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nDear authors,\nI appreciated your manuscript very much.\nCongratulations for all the details that you presented in the work. I had some suggestions, regarding the sample that was analyzed in the work, considering that the majority of patients in the sample belong to classes I or II of the NYHA, and as you declare in the limitations, “more than 90% of our patients are within the low risk group of 0-10%”. Your work could have the advantage to be done in one single service, with the same quality of the surgery. Unfortunately, you said that the performance of the surgeons is not the same, between them.\nParticularly, I have critiques about the multicenter studies, which have a lot of qualities, but the surgeons from different services and nations have different results, like you expressed in the limitations. The problem of the work being retrospective brings some problems, but this does not compromise the results.\nYour discussion was very clear, adding your article in the present context of the literature. The conclusions are simple and straightforward, allowing the Cardiac Surgeons of Malaysia to use the EuroSCORE II, securely in your country.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-534
https://f1000research.com/articles/8-59/v1	15 Jan 19	{ "type": "Research Article", "title": "Factors associated with poor physical performance in older adults of 11 Peruvian high Andean communities", "authors": [ "Diego Urrunaga-Pastor", "Fernando M. Runzer-Colmenares", "Tania M. Arones", "Rosario Meza-Cordero", "Silvana Taipe-Guizado", "Jack M. Guralnik", "Jose F. Parodi", "Diego Urrunaga-Pastor", "Tania M. Arones", "Rosario Meza-Cordero", "Silvana Taipe-Guizado", "Jack M. Guralnik", "Jose F. Parodi" ], "abstract": "Background: Physical performance in the older adult has been extensively studied. However, only a few studies have evaluated physical performance among older adults of high Andean populations and none have studied the factors associated with it. The objective of this study was to evaluate factors associated with poor physical performance by using the Short Physical Performance Battery (SPPB) in older adults living in 11 Peruvian high Andean communities. Methods: An analytical cross-sectional study was carried out in inhabitants aged 60 or over from 11 high-altitude Andean communities of Peru during 2013-2017. Participants were categorized in two groups according to their SPPB score: poor physical performance (0-6 points) and medium/good physical performance (7-12 points). Additionally, we collected socio-demographic, medical, functional and cognitive assessment information. Poisson regression models were constructed to identify factors associated with poor physical performance. Prevalence ratio (PR) with 95% confidence intervals (95 CI%) are presented. Results: A total of 407 older adults were studied. The average age was 73.0 ± 6.9 years (range: 60-94 years) and 181 (44.5%) participants had poor physical performance (0-6 points). In the adjusted Poisson regression analysis, the factors associated with poor physical performance were: female gender (PR=1.29; 95%CI: 1.03-1.61), lack of social support (PR=2.10; 95%CI: 1.17-3.76), number of drugs used (PR=1.09; 95%CI: 1.01-1.17), urinary incontinence (PR=1.45; 95%CI: 1.16-1.82), exhaustion (PR=1.35; 95%CI: 1.03-1.75) and cognitive impairment (PR=1.89; 95%CI: 1.40-2.55). Conclusions: Almost half of the population evaluated had poor physical performance based on the SPPB. Factors that would increase the possibility of suffering from poor physical performance were: female gender, lack of social support, number of drugs used, urinary incontinence, exhaustion and cognitive impairment. Future studies with a larger sample and longitudinal follow-up are needed to design beneficial interventions for the high Andean population.", "keywords": [ "Physical performance", "Altitude", "Elderly", "Latin America", "Peru" ], "content": "Introduction\n\nAging is a physiological process that involves changes in respiratory, cardiovascular, muscular, kidney and brain function1–4. In addition, these changes organically could be exacerbated in older adults living at high altitude due to the hypoxia to which they are chronically exposed, increasing their risk of suffering certain pathologies; however, there is no consensus surrounding this situation5,6. Chronic mountain sickness is a clinical syndrome that affects natives or residents living for a long time at an altitude greater than 2500 meters above sea level (masl) and is characterized by erythrocytosis that could evolve to severe pulmonary hypertension and generate congestive heart failure, affecting the ability of Andean older adults to maintain their daily activities and their physical performance7.\n\nPhysical performance in the older adult has been extensively studied, and poor nutritional status8, sarcopenia9, decreased muscle mass, frailty10, sarcopenic obesity11, mortality12, disability13 and dementia14, common chronic diseases of aging, have been associated. A previous study conducted in rural Peruvian communities located at 3345 and 6 masl found that the prevalence of poor physical performance in older adults living in rural communities at sea level was twice as high as that of older adults that are residing in rural areas at high altitude15.\n\nPrevious studies in high-altitude communities have described older population´s nutritional status, finding a prevalence rates of 9.4% for malnutrition16, 17.6% for sarcopenia17, 15.2% for insomnia18, 12.2% for frailty19 and 75.2% for fear of falling19. These figures are similar to those described in populations at sea level20–23. At high altitudes, an increased ventilatory response and a lower cardiac response to hypoxia will favor oxygen uptake in the lungs and allow the maintenance of a normal oxygen saturation, at moderate exercise24. In addition, tissue hypoxia, oxidative stress and the action of free radicals would be increased, affecting cardiac energy metabolism and skeletal muscle performance; in this way, a decrease in mitochondrial volume would be generated. This situation would occur in people exposed for a long time or who have returned from high altitude25,26 and significantly affect the physical performance of the older adult living at high altitude; however, there is no clear consensus regarding this process26–28, especially in the elderly.\n\nAdditionally, there are no parameters or determinants in relation to the poor physical performance in older adults of the Andes, which could be different from those described for other populations, due to social or geographical conditions, or due to access to health services. Therefore, this study aimed to determine the factors associated with poor physical performance in older adults from 11 high Andean communities in Peru.\n\n\nMethods\n\nAnalytical cross-sectional study, carried out in inhabitants aged 60 or over from 11 high-altitude (≥1500 masl)29 Andean communities of Peru: La Jalca, Leimebamba (Amazonas), Llupa, San Pedro de Chaná, Atipayan (Áncash), Pampamarca (Huánuco), Chacapampa (Huancayo), Ayahuanco (Ayacucho), Paucarcolla (Puno), Vilca (Huancavelica) and Viñac (Lima) during 2013–2017 period. All inhabitants of the 11 high-altitude Andean communities included, belonged to the same ethnic group and performed a similar work activity, based mainly on agriculture, farming and trading30.\n\nThe National Statistics Institute of Peru (Instituto Nacional de Estadística e Informática -INEI) classifies communities with 100 houses not in a capital district, that have more than 100 individuals, located in a dispersed way without forming blocks as rural communities31. The communities were located in the Peruvian highlands as follows: a) La Jalca: urban settlement located at 2800 masl; b) Leimebamba: rural village located at 2158 masl; c) Llupa: rural village located at 3511 masl; d) San Pedro de Chaná: rural village located at 3413 masl; e) Atipayán: rural village located at 3364 masl; f) Pampamarca: urban village located at 3445 masl; g) Chacapampa: rural village located at 3358 masl; h) Ayahuanco: rural village located at 3414 masl; i) Paucarcolla: urban village located at 3847 masl; j) Vilca: rural village located at 3275 masl; k) Viñac: rural village located at 3315 masl. The Peruvian Andes weather biodiversity includes high temperatures, rainfall and cloudy seasons32. These areas have low levels of pollution; however, mining activities are endangering ecosystems sustainability33.\n\nA non-probabilistic, census-type sampling was performed, registering all the elderly people in the highland communities previously described. We included all or most (approximately 95%) of the geriatric population of each community (urban/rural)34. The analysis unit was elderly person from high-altitude Andean communities (rural/urban). The final sample included 413 older adults who voluntarily signed an informed consent form accepting their participation in the study.\n\nParticipants were visited in their homes up to three times to be invited to participate in the study. Those who agreed to participate voluntarily signed a document of informed consent prior to the collection of data by the researchers34. Data was collected on sociodemographic characteristics, medical background (falls, polypharmacy, comorbidities, tobacco, alcohol and coca leaf consumption), Barthel Index, Edmonton test, exhaustion)35–37, physical performance (Short Physical Performance Battery)12, anthropometric measurements (height and weight) and cognitive status (Yesavage test and Pfeiffer Questionnaire)38,39. The interview was conducted by a geriatrician, medical doctors and medical students (previously trained by the geriatrician). All the self-reported data was collected during the interview.\n\nOutcome: Poor physical performance. To evaluate physical performance in the participants, we used the Short Physical Performance Battery (SPPB). The SPPB is based on three timed tasks: standing balance, walking or gait speed, and five repetitive chair stands. The timed results of each subtest are rescaled according to predefined cut points for obtaining a score ranging from 0 (worst performance) to 12 (best performance)40. The variable was categorized as: poor physical performance (0-6) and medium/good physical performance (7-12)12,15.\n\nOther variables\n\nSociodemographic characteristics. The sociodemographic characteristics included and evaluated by self-report were: age (less than or equal to 70 years, 71 to 80 years, over 80 years), gender (male, female), educational level (no education/incomplete elemental school, complete elemental school, complete high school), marital status (single, married, widowed/divorced), live alone (yes or no), time by foot from their home to the nearest health centre (in minutes) and altitude (masl). The sociodemographic information was corroborated with the participant's national identity document (ID card).\n\nMedical background. The following variables were included and evaluated by self-report: falls in the last year (none, at least 1), hospitalizations in the last year (none, at least 1), polypharmacy (5 drugs or more, under medical prescription)41, tobacco consumption (yes or no), alcohol consumption (yes or no), coca leaf consumption (yes or no), high blood pressure (HBP) (yes or no), diabetes mellitus type 2 (DM2) (yes or no), chronic obstructive pulmonary disease (COPD) (yes or no) and low back pain (yes or no). Likewise, a variable of comorbidities (obesity defined according to body mass index (BMI) + HBP + COPD + DM2 + low back pain) was constructed34,42. The medical background information was confirmed by the caregiver/family member at the time of data collection.\n\nWe determined the body mass index (BMI), which was calculated with the formula weight in kg/(size in meters squared). This was categorized as follows: malnutrition (<18.5 kg/m2), normal (18.5-24.99 kg/m2), overweight (25.0-29.99 kg/m2) and obesity (>30.0 kg/m2)43.\n\nFunctional assessment. We used the Barthel Index, a questionnaire about 10 basic activities of daily living (ADL) with a total score between 0–100. It was analyzed as a continuous variable and also divided into two strata: independent (100) and dependent (<100)35,44.\n\nAdditionally, we use two items from the Edmonton test: 1) social support: When you need help, do you have someone who meets your needs? (always, sometimes/never); 2) urinary incontinence: Do you have trouble holding urine when you do not feel like urinating? (yes or no). The Edmonton test has 9 items and is used to evaluate frailty36.\n\nIn the present study, we evaluated exhaustion, which is defined by 3 items that the participant must respond according to the way he felt during the last 2 weeks: 1) did you feel full of energy? (yes or no); 2) did you feel that you could not go on? (yes or no); 3) did you feel that all you did was with effort? (yes or no). A score equal or greater than two was considered positive for exhaustion dimension37,45.\n\nPsychological and cognitive assessment. We used the Yesavage test, which is a 5-item questionnaire that evaluates the presence of depressive symptoms. A score equal or greater than three was considered positive for depressive symptoms38.\n\nWe used the Pfeiffer Questionnaire, a 10-item questionnaire for evaluation of cognitive impairment. The strata were generated as follows: no impairment (0 to 2 errors), mild impairment (3 to 4 errors), moderate impairment (5-7 errors)39.\n\nWe used STATA v14.0 for our analysis. Descriptive results were presented using measures of central tendency, dispersion measures, absolute frequencies, and relative frequencies. The characteristics of the participants with poor and medium/good physical performance were compared using the Chi square test, Fisher’s exact test, Student’s T test or the Wilcoxon rank sum test as appropriate.\n\nTwo Poisson regression models (1 crude and 1 adjusted) were constructed using robust variance with the objective of evaluating factors associated with poor physical performance in the participants. The reported measure was the prevalence ratio (PR) with their respective 95% confidence intervals (95%CI).\n\nThe adjusted model included the following variables: gender, lack of social support, alcohol consumption, tobacco consumption, number of drugs used, comorbidities, urinary incontinence, falls in the last year, hospitalizations in the last year, dependence ADL, exhaustion, depressive symptoms, exhaustion, cognitive impairment and altitude (masl). These variables were included in the adjusted model because they had statistically significant association with poor physical performance in the crude Poisson regression analysis. Additionally, we evaluated the possible collinearity between the exposure variables entered in the adjusted model.\n\nThe research project was approved by the Institutional Review Board of the Peruvian Naval Medical Centre, located in Lima, Peru. Informed consent was obtained from all the participants. In case of cognitive impairment, the family member who was present at the time of data collection gave the written consent. Furthermore, the anonymity of the participants and confidentiality of the data were ensured.\n\n\nResults\n\nOf a total of 413 elderly adults, 3 participants were excluded because of severe cognitive impairment, equivalent to a score equal or greater than 8 in the Pfeiffer Questionnaire, 2 participants were excluded because they did not have variables of interest and 1 participant was excluded because of being physically incapable of performing the physical and functional performance tests (visual and auditory impairment). Finally, a total of 407 individuals were analyzed.\n\nData from 407 elderly adults from 11 high Andean communities were analyzed. In total, 181 (44.5%) participants had poor physical performance and the SPPB mean was 7.3 ± 3.1. The mean age was 73.0 ± 6.9 years old (range: 60–94 years old), 267 (65.6%) participants were female, 335 (82.3%) did not count with education or had not finished elementary school, 271 (77.2%) worked in agriculture and 91 (22.4%) lived alone. Statistically significant differences were found in gender, educational level, live alone, time by foot from their home to the nearest health centre (in minutes) and altitude (masl) among physical performance groups (Table 1). Full raw data are available on OSF46.\n\nMean ± standard deviation. Median (interquartile range).\n\nOf the 407 elderly adults evaluated, 261 (64.3%) had at least 1 fall in the last year, 48 (11.8%) were hospitalized at least once in the last year, 74 (18.2%) consumed coca leaf, 109 (19.4%) were obese according to BMI, 337 (83.0%) had disability (Barthel Index), 150 (36.9%) had depressive symptoms and 116 (28.5%) had cognitive impairment (mild-moderate) (Table 2).\n\nMedian (interquartile range). HBP, high blood pressure; COPD, chronic obstructive pulmonary disease. DM2, diabetes mellitus type 2; BMI, body mass index.\n\nIn the adjusted Poisson regression analysis, the factors associated with poor physical performance were: female gender (PR=1.29; 95%CI: 1.03-1.61), lack of social support (PR=2.10; 95%CI: 1.17-3.76), number of drugs used (PR=1.09; 95%CI: 1.01-1.17), urinary incontinence (PR=1.45; 95%CI: 1.16-1.82), exhaustion (PR=1.35; 95%CI: 1.03-1.75) and cognitive impairment (PR=1.89; 95%CI: 1.40-2.55) (Table 3).\n\n1Activities of daily living, assessed with Barthel Index. 2Altitude for each 1000 masl.\n\n\nDiscussion\n\nA total of 407 older adults from 11 high Andean communities were analyzed, of whom 44.5% had poor physical performance. Factors that would increase the possibility of suffering poor physical performance were: female gender, lack of social support, number of drugs used, urinary incontinence, exhaustion and cognitive impairment.\n\nPrevious studies evaluated physical performance using SPPB as a measurement tool, finding diverse results. One of them, conducted in the United States with 631 older adults, calculated a SPPB average score of 9.9 in its participants, being higher than the 7.3 points found as a mean score in our population. However, the sample size of older adults lived in urban areas and was higher than the one we assessed47. On the other hand, in the InCHIANTI study cohort conducted in 542 older adults from Italy, it was found that approximately 65% of the participants with a SPPB score less than or equal to 7 were unable to complete the 400 meters walk test after the three years of follow-up, being a higher proportion than the found in our study. In addition, this SPPB score (≤7) was associated with an odds ratio (OR) of approximately 27 predicting inability to complete 400 meters walk test in those able to walk 400 meters at baseline48. Similarly, another study conducted in Italy found a lower SPPB mean score than the one calculated in our study population; nevertheless, this study was performed in hospitalized patients49. Additionally, a previous study carried out in Peru in the rural communities of Atipayán (3345 masl) and Santa (6 masl), showed a prevalence of poor physical performance of 10.0% and 19.4%, respectively, both lower figures to that found in this study15.\n\nWe have found a higher SPPB mean score than that found in other studies, highlighting the fact of being a population living in altitude cities. Nevertheless, these findings can only be interpreted for the altitude ranges evaluated in the present study.\n\nWe found an association between female gender and poor physical performance in the evaluated population. Equally, a cohort carried out in 3041 well-functioning white and black men and women, aged 70–79 years, found that men independently of the race had a better physical performance than women (evaluated by the knee extension strength, chair-rise, 6 meters walk time, 400 meters walk time and standing balance test)50. In contrast, Vasunilashorn et al.48 did not find differences between physical performance groups and gender. This association could be explained because women usually have less muscle mass than men, and menopause produce an acute decline in strength and muscle mass, compared with the gradual loss of strength by men of similar age17,51.\n\nIn this study, association between lack of social support and poor physical performance was found. A systematic review by Vagetti et al. during 2014 that aimed to assess the association between physical activity and quality of life in older adults found a moderate association between social support and physical activity in older adults52. Similarly, a study in Norway found a consistent correlation between physical activity in older adults and social support, especially regarding family social support rather than friend-related support53. This association would be explained by the close relationship between the deterioration of physical and mental health caused by the lack of social support in older adults, which would negatively affect the control of diseases and the physical performance of this population54.\n\nThe presence of chronic diseases and comorbidities are common in the older people, and require pharmacological therapy in the majority of cases in order to manage them properly55. A study conducted in 1123 hospitalized older adults in Italy found that the prevalence of polypharmacy was higher in patients with poor physical performance and grip strength56. Also, the association between consumption of more than five drugs would be associated with the presence of frailty, disability and falls in older adults, which would significantly affect the physical performance of the elderly57,58. Due to the absence of an adequate health network in high-altitude areas able to properly provide drugs to older people59,60, the presence of polypharmacy would be significantly lower than that of the older people in urban areas, limiting the consequences in their physical performance.\n\nWe found an association between urinary incontinence and poor physical performance in the population that was evaluated. A study conducted in Taiwan by Chiu et al. found an association between poor physical performance and the presence of urinary incontinence in older adults61. Similarly, in a cohort study conducted in 328 older Latinos in the United States, the increase in SPPB score at one-year follow-up was associated with a lower incidence of urinary incontinence62.\n\nIn this study, an association between exhaustion and poor physical performance was found. Exhaustion and poor physical performance evaluated by SPPB are useful tools in the evaluation of sarcopenia, frailty and disability63–66. Previous studies reinforce the association found in this study, describing very low SPPB scores in fragile older people compared to non-fragile older people (2.9 vs. 8.5, respectively)67. In our study population, a high prevalence of exhaustion was found, which could be due to the continuous physical effort that these inhabitants perform in their daily activities, which mainly involve agriculture and trading.\n\nWe found no association between poor physical performance and disability. As well as SPPB, the functional reach test, both performance-based measure, was not associated with disability assessed by the Barthel Index in older adults of Peruvian high Andean communities34. In addition, we did not find an association between poor physical performance and altitude in the adjusted regression model. Both associations had statistical significance in the crude regression model; however, in the adjusted model, they lost it. A possible explanation for this could be the sample size, because, in the adjusted model, both associations presented a p-value with marginal significance68. Although p-value is a useful parameter to explain a result based on statistical significance, it is not the only one to be taken into account69.\n\nThe relevance of our results allows our research team to hypothesize plausible explanations of the presented findings: 1) people with a high number of comorbidities cannot live at highest altitudes, so we do not find a comorbid population in our study; 2) living at that altitude range makes you physically stronger; 3) there is another variable or condition about the people living at high altitude that was missed in our study and that we did not adjust for in the regression models. In regard of these, the Andean older people work from a very young age in tasks that involve physical effort, so this could be an interesting point of the study. It is also important to indicate that in the crude model, altitude (for each 1000 masl) increased the probability of poor physical performance; however, after we adjusted the analysis including medical, functional and cognitive variables, the high altitude became a protective marker for poor physical function. These questions would serve as a basis for future studies.\n\nMoreover, an association between the presence of cognitive impairment and poor physical performance was found. The protective effect of physical activity against the development of some type of dementia or neurocognitive disorder has been previously described in multiple studies70–72. In rural populations at sea level and in altitude, the prevalence of cognitive disorders is low; this could be attributed to different lifestyles, such as the constant physical activity they have performed throughout their lives15.\n\nThis study has some limitations: 1) the sampling conducted was not probabilistic, the results cannot be extrapolated; nevertheless, this study was conducted in 11 communities at different altitudes, and the participants reported fewer comorbidities than persons in hospitals, drawing closer to the rural reality; 2) because of its cross-sectional design, this study does not allow us to evaluate causality between the poor physical performance and the associated factors; yet, we still could identify useful markers for future intervention studies; 3) we used self-report to collect some variables in this study which can generate a recall bias. Nevertheless, this is not the case of our main variable which was performance-based measured73; 4) low educational level of the studied population would affect the accuracy of self-report to collect information on complex diseases42; hence, we corroborated the data of the most common comorbidities with a family member/caregiver of the respondent at the time of the interview; 5) because of their low educational level, it was not possible to assess the amount of alcohol of tobacco consumed by the participants; 6) some variables studied have missing values, though, they did not exceed 20%, allowing its analysis74.\n\nIn conclusion, almost half of the population evaluated had poor physical performance based on the SPPB. Factors that would increase the possibility of suffering from poor physical performance were: female gender, lack of social support, number of drugs used, urinary incontinence, exhaustion and cognitive impairment. These markers would be very important to develop future cohort studies which would like to study more specifically some marker found in this study.\n\n\nData availability\n\nThe raw data associated with this study are available on OSF. DOI: https://doi.org/10.17605/OSF.IO/RSC7Q46.\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nWe acknowledge the staff of the Aging Investigation Center - Faculty of Medicine at the Universidad de San Martín de Porres, Peru; and the staff of Geriatric Service of the Peruvian Naval Medical Center for the logistical support provided.\n\n\nReferences\n\nClegg A, Young J, Iliffe S, et al.: Frailty in elderly people. Lancet. 2013; 381(9868): 752–62. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJanssens JP, Pache JC, Nicod LP: Physiological changes in respiratory function associated with ageing. Eur Respir J. 1999; 13(1): 197–205. PubMed Abstract\n\nWeinstein JR, Anderson S: The aging kidney: physiological changes. Adv Chronic Kidney Dis. 2010; 17(4): 302–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPugh KG, Wei JY: Clinical implications of physiological changes in the aging heart. Drugs Aging. 2001; 18(4): 263–76. PubMed Abstract \| Publisher Full Text\n\nLipsitz LA: Dynamics of stability: the physiologic basis of functional health and frailty. J Gerontol A Biol Sci Med Sci. 2002; 57(3): B115–B25. PubMed Abstract \| Publisher Full Text\n\nRodway GW, Hoffman LA, Sanders MH: High-altitude-related disorders--Part II: prevention, special populations, and chronic medical conditions. Heart Lung. 2004; 33(1): 3–12. PubMed Abstract \| Publisher Full Text\n\nLeón-Velarde F, Maggiorini M, Reeves JT, et al.: Consensus statement on chronic and subacute high altitude diseases. High Alt Med Biol. 2005; 6(2): 147–57. PubMed Abstract \| Publisher Full Text\n\nPenninx BW, Pahor M, Cesari M, et al.: Anemia is associated with disability and decreased physical performance and muscle strength in the elderly. J Am Geriatr Soc. 2004; 52(5): 719–24. PubMed Abstract \| Publisher Full Text\n\nJanssen I: Influence of sarcopenia on the development of physical disability: the Cardiovascular Health Study. J Am Geriatr Soc. 2006; 54(1): 56–62. PubMed Abstract \| Publisher Full Text\n\nLandi F, Onder G, Russo A, et al.: Calf circumference, frailty and physical performance among older adults living in the community. Clin Nutr. 2014; 33(3): 539–44. PubMed Abstract \| Publisher Full Text\n\nRolland Y, Lauwers-Cances V, Cristini C, et al.: Difficulties with physical function associated with obesity, sarcopenia, and sarcopenic-obesity in community-dwelling elderly women: the EPIDOS (EPIDemiologie de l'OSteoporose) Study. Am J Clin Nutr. 2009; 89(6): 1895–900. PubMed Abstract \| Publisher Full Text\n\nGuralnik JM, Simonsick EM, Ferrucci L, et al.: A short physical performance battery assessing lower extremity function: association with self-reported disability and prediction of mortality and nursing home admission. J Gerontol. 1994; 49(2): M85–M94. PubMed Abstract \| Publisher Full Text\n\nGuralnik JM, Ferrucci L, Simonsick EM, et al.: Lower-extremity function in persons over the age of 70 years as a predictor of subsequent disability. N Engl J Med. 1995; 332(9): 556–61. PubMed Abstract \| Publisher Full Text\n\nWang L, Larson EB, Bowen JD, et al.: Performance-based physical function and future dementia in older people. Arch Intern Med. 2006; 166(10): 1115–20. PubMed Abstract \| Publisher Full Text\n\nEstela-Ayamamani D, Espinoza-Figueroa J, Columbus-Morales M, et al.: [Physical performance of older adults living in rural areas at sea level and at high altitude in Peru]. Rev Esp Geriatr Gerontol. 2015; 50(2): 56–61. PubMed Abstract \| Publisher Full Text\n\nTramontano A, Veronese N, Giantin V, et al.: Nutritional status, physical performance and disability in the elderly of the Peruvian Andes. Aging Clin Exp Res. 2016; 28(6): 1195–201. PubMed Abstract \| Publisher Full Text\n\nTramontano A, Veronese N, Sergi G, et al.: Prevalence of sarcopenia and associated factors in the healthy older adults of the Peruvian Andes. Arch Gerontol Geriatr. 2017; 68: 49–54. PubMed Abstract \| Publisher Full Text\n\nSakamoto R, Okumiya K, Norboo T, et al.: Sleep quality among elderly high-altitude dwellers in Ladakh. Psychiatry Res. 2017; 249: 51–57. PubMed Abstract \| Publisher Full Text\n\nCurcio CL, Henao GM, Gomez F: Frailty among rural elderly adults. BMC Geriatr. 2014; 14(1): 2. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDa Mata FA, Pereira PP, Andrade KR, et al.: Prevalence of Frailty in Latin America and the Caribbean: A Systematic Review and Meta-Analysis. PLoS One. 2016; 11(8): e0160019. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDíaz-Villegas G, Parodi J, Merino-Taboada A, et al.: Calf circumference and risk of falls among Peruvian older adults. Eur Geriatr Med. 2016; 7(6): 543–6. Publisher Full Text\n\nEthgen O, Beaudart C, Buckinx F, et al.: The Future Prevalence of Sarcopenia in Europe: A Claim for Public Health Action. Calcif Tissue Int. 2017; 100(3): 229–234. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKalaiselvi S, Arjumand Y, Jayalakshmy R, et al.: Prevalence of under-nutrition, associated factors and perceived nutritional status among elderly in a rural area of Puducherry, South India. Arch Gerontol Geriatr. 2016; 65: 156–60. PubMed Abstract \| Publisher Full Text\n\nRichalet JP, Lhuissier FJ: Aging, Tolerance to High Altitude, and Cardiorespiratory Response to Hypoxia. High Alt Med Biol. 2015; 16(2): 117–24. PubMed Abstract \| Publisher Full Text\n\nMurray AJ: Energy metabolism and the high-altitude environment. Exp Physiol. 2016; 101(1): 23–7. PubMed Abstract \| Publisher Full Text\n\nHoppeler H, Mueller M, Vogt M: Skeletal muscle tissue changes with hypoxia. High Altitude. Springer; 2014; 191–202. Publisher Full Text\n\nGilbert-Kawai ET, Milledge JS, Grocott MP, et al.: King of the mountains: Tibetan and Sherpa physiological adaptations for life at high altitude. Physiology (Bethesda). 2014; 29(6): 388–402. PubMed Abstract \| Publisher Full Text\n\nHoppeler H, Vogt M: Muscle tissue adaptations to hypoxia. J Exp Biol. 2001; 204(Pt 18): 3133–9. PubMed Abstract\n\nTaylor AT: High-altitude illnesses: physiology, risk factors, prevention, and treatment. Rambam Maimonides Med J. 2011; 2(1): e0022. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMiranda JJ, Gilman RH, Smeeth L: Differences in cardiovascular risk factors in rural, urban and rural-to-urban migrants in Peru. Heart. 2011; 97(10): 787–96. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSociales INdEeIDTdDeI: Perfil sociodemográfico del Perú: Censos Nacionales 2007: XI de población y VI de vivienda. INEI; 2008. Reference Source\n\nHerzog SK, Martínez R, Jørgensen PM, et al.: Climate change and biodiversity in the tropical Andes. Inter-American Institute for Global Change Research (IAI) and Scientific Committee on Problems of the Environment (SCOPE). 2011. Publisher Full Text\n\nRolando JL, Turin C, Ramírez DA, et al.: Key ecosystem services and ecological intensification of agriculture in the tropical high-Andean Puna as affected by land-use and climate changes. Agric Ecosyst Environ. 2017; 236: 221–33. Publisher Full Text\n\nUrrunaga-Pastor D, Moncada-Mapelli E, Runzer-Colmenares FM, et al.: Factors associated with poor balance ability in older adults of nine high-altitude communities. Arch Gerontol Geriatr. 2018; 77: 108–14. PubMed Abstract \| Publisher Full Text\n\nCollin C, Wade DT, Davies S, et al.: The Barthel ADL Index: a reliability study. Int Disabil Stud. 1988; 10(2): 61–3. PubMed Abstract \| Publisher Full Text\n\nRolfson DB, Majumdar SR, Tsuyuki RT, et al.: Validity and reliability of the Edmonton Frail Scale. Age Ageing. 2006; 35(5): 526–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFried LP, Tangen CM, Walston J, et al.: Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci. 2001; 56(3): M146–56. PubMed Abstract \| Publisher Full Text\n\nYesavage JA, Sheikh JI: 9/Geriatric depression scale (GDS) recent evidence and development of a shorter version. Clin Gerontol. 1986; 5(1–2): 165–73. Publisher Full Text\n\nPfeiffer E: A short portable mental status questionnaire for the assessment of organic brain deficit in elderly patients. J Am Geriatr Soc. 1975; 23(10): 433–41. PubMed Abstract \| Publisher Full Text\n\nPavasini R, Guralnik J, Brown JC, et al.: Short Physical Performance Battery and all-cause mortality: systematic review and meta-analysis. BMC Med. 2016; 14(1): 215. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nViktil KK, Blix HS, Moger TA, et al.: Polypharmacy as commonly defined is an indicator of limited value in the assessment of drug-related problems. Br J Clin Pharmacol. 2007; 63(2): 187–95. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOkura Y, Urban LH, Mahoney DW, et al.: Agreement between self-report questionnaires and medical record data was substantial for diabetes, hypertension, myocardial infarction and stroke but not for heart failure. J Clin Epidemiol. 2004; 57(10): 1096–103. PubMed Abstract \| Publisher Full Text\n\nFlegal KM, Carroll MD, Ogden CL, et al.: Prevalence and trends in obesity among US adults, 1999-2000. JAMA. 2002; 288(14): 1723–7. PubMed Abstract \| Publisher Full Text\n\nRunzer-Colmenares FM, Samper-Ternent R, Al Snih S, et al.: Prevalence and factors associated with frailty among Peruvian older adults. Arch Gerontol Geriatr. 2014; 58(1): 69–73. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNascimento PPP, Batistoni SST, Neri AL: Frailty and depressive symptoms in older adults: data from the FIBRA study-UNICAMP. Psicol Reflex Crit. 2016; 29(1): 16. Publisher Full Text\n\nPastor DU: Factors Associated to Poor Physical Performance in Older Adults of 11 Peruvian High Andean Communities. OSF. Web. 2019.\n\nVerghese J, Holtzer R, Lipton RB, et al.: Mobility stress test approach to predicting frailty, disability, and mortality in high-functioning older adults. J Am Geriatr Soc. 2012; 60(10): 1901–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVasunilashorn S, Coppin AK, Patel KV, et al.: Use of the Short Physical Performance Battery Score to predict loss of ability to walk 400 meters: analysis from the InCHIANTI study. J Gerontol A Biol Sci Med Sci. 2009; 64(2): 223–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVolpato S, Cavalieri M, Sioulis F, et al.: Predictive value of the Short Physical Performance Battery following hospitalization in older patients. J Gerontol A Biol Sci Med Sci. 2011; 66(1): 89–96. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTaaffe DR, Simonsick EM, Visser M, et al.: Lower extremity physical performance and hip bone mineral density in elderly black and white men and women: cross-sectional associations in the Health ABC Study. J Gerontol A Biol Sci Med Sci. 2003; 58(10): M934–M42. PubMed Abstract \| Publisher Full Text\n\nHughes VA, Frontera WR, Roubenoff R, et al.: Longitudinal changes in body composition in older men and women: role of body weight change and physical activity. Am J Clin Nutr. 2002; 76(2): 473–81. PubMed Abstract \| Publisher Full Text\n\nVagetti GC, Barbosa Filho VC, Moreira NB, et al.: Association between physical activity and quality of life in the elderly: a systematic review, 2000-2012. Braz J Psychiatry. 2014; 36(1): 76–88. PubMed Abstract \| Publisher Full Text\n\nHansen BH, Ommundsen Y, Holme I, et al.: Correlates of objectively measured physical activity in adults and older people: a cross-sectional study of population-based sample of adults and older people living in Norway. Int J Public Health. 2014; 59(2): 221–30. PubMed Abstract \| Publisher Full Text\n\nLiu L, Gou Z, Zuo J: Social support mediates loneliness and depression in elderly people. J Health Psychol. 2016; 21(5): 750–8. PubMed Abstract \| Publisher Full Text\n\nFries JF: Aging, natural death, and the compression of morbidity. 1980. Bull World Health Organ. 2002; 80(3): 245–50. PubMed Abstract \| Free Full Text\n\nSganga F, Vetrano DL, Volpato S, et al.: Physical performance measures and polypharmacy among hospitalized older adults: results from the CRIME study. J Nutr Health Aging. 2014; 18(6): 616–21. PubMed Abstract \| Publisher Full Text\n\nGnjidic D, Hilmer SN, Blyth FM, et al.: Polypharmacy cutoff and outcomes: five or more medicines were used to identify community-dwelling older men at risk of different adverse outcomes. J Clin Epidemiol. 2012; 65(9): 989–95. PubMed Abstract \| Publisher Full Text\n\nDelbaere K, Van den Noortgate N, Bourgois J, et al.: The Physical Performance Test as a predictor of frequent fallers: a prospective community-based cohort study. Clin Rehabil. 2006; 20(1): 83–90. PubMed Abstract \| Publisher Full Text\n\nLiu X, Gao W, Yan H: Measuring and decomposing the inequality of maternal health services utilization in western rural China. BMC Health Serv Res. 2014; 14(1): 102. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMinisterio de Salud, Dirección General de Gestión del Desarrollo de Recursos Humanos: Segunda medición de las metas regionales de recursos humanos para la salud: Perú 2007-2015. Minsa Lima; 2013. Reference Source\n\nChiu AF, Huang MH, Hsu MH, et al.: Association of urinary incontinence with impaired functional status among older people living in a long-term care setting. Geriatr Gerontol Int. 2015; 15(3): 296–301. PubMed Abstract \| Publisher Full Text\n\nMorrisroe SN, Rodriguez LV, Wang PC, et al.: Correlates of 1-year incidence of urinary incontinence in older Latino adults enrolled in a community-based physical activity trial. J Am Geriatr Soc. 2014; 62(4): 740–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIolascon G, Di Pietro G, Gimigliano F, et al.: Physical exercise and sarcopenia in older people: position paper of the Italian Society of Orthopaedics and Medicine (OrtoMed). Clin Cases Miner Bone Metab. 2014; 11(3): 215–21. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCooper C, Dere W, Evans W, et al.: Frailty and sarcopenia: definitions and outcome parameters. Osteoporos Int. 2012; 23(7): 1839–48. PubMed Abstract \| Publisher Full Text\n\nBrown M, Sinacore DR, Binder EF, et al.: Physical and performance measures for the identification of mild to moderate frailty. J Gerontol A Biol Sci Med Sci. 2000; 55(6): M350–M5. PubMed Abstract \| Publisher Full Text\n\nVestergaard S, Nayfield SG, Patel KV, et al.: Fatigue in a representative population of older persons and its association with functional impairment, functional limitation, and disability. J Gerontol A Biol Sci Med Sci. 2009; 64(1): 76–82. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBuckinx F, Reginster JY, Petermans J, et al.: Relationship between frailty, physical performance and quality of life among nursing home residents: the SENIOR cohort. Aging Clin Exp Res. 2016; 28(6): 1149–57. PubMed Abstract \| Publisher Full Text\n\nAi C, Norton EC: Interaction terms in logit and probit models. Econ Lett. 2003; 80(1): 123–9. Publisher Full Text\n\nGreenland S, Senn SJ, Rothman KJ, et al.: Statistical tests, P values, confidence intervals, and power: a guide to misinterpretations. Eur J Epidemiol. 2016; 31(4): 337–50. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLaurin D, Verreault R, Lindsay J, et al.: Physical activity and risk of cognitive impairment and dementia in elderly persons. Arch Neurol. 2001; 58(3): 498–504. PubMed Abstract \| Publisher Full Text\n\nHeyn P, Abreu BC, Ottenbacher KJ: The effects of exercise training on elderly persons with cognitive impairment and dementia: a meta-analysis. Arch Phys Med Rehabil. 2004; 85(10): 1694–704. PubMed Abstract \| Publisher Full Text\n\nMiddleton LE, Barnes DE, Lui LY, et al.: Physical activity over the life course and its association with cognitive performance and impairment in old age. J Am Geriatr Soc. 2010; 58(7): 1322–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nReuben DB, Siu AL, Kimpau S: The predictive validity of self-report and performance-based measures of function and health. J Gerontol. 1992; 47(4): M106–M10. PubMed Abstract \| Publisher Full Text\n\nDong Y, Peng CY: Principled missing data methods for researchers. SpringerPlus. 2013; 2(1): 222. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "43043", "date": "21 Feb 2019", "name": "Oscar Rosas-Carrasco", "expertise": [ "Reviewer Expertise Geriatrics", "Body composition", "sarcopenia", "frailty" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis work has a clear objective. The lack of studies in communities far from the big cities will always be a solid point to support the publication of these studies. The strength of the study is that it includes the possibility of studying the physical performance in populations with height above sea level greater than 2500 meters.\nI have some considerations to correct in the methodology:\nI understand that the authors preferred to use a Poisson regression due to the fact that the characteristics of the dependent variable (physical performance) were adequate (by distribution?) to prefer this analysis regression, however it should be specifically noted and included why they did not use logistic regression if the dependent variable presented only had two categories. On the other hand, include if medical doctors and medical students were compared with any statistical test to corroborate concordance. Do the authors have information about the migration of those with high comorbidity? The above could explain why a low frequency of chronic diseases and good performance were found. Include some result of the final Poisson regression model that allows to know if the fit of model regression were adequate to present the results obtained.\nIn conclusion, the manuscript must be accepted with some corrections that the authors must consider, the results are interesting and provide novel information in these regions of Peru.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "43042", "date": "01 Mar 2019", "name": "Diego Andrés Chavarro-Carvajal", "expertise": [ "Reviewer Expertise Geriatrics", "nutrition", "dementia", "frailty", "sarcopenia." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nUrrunaga-Pastor and co-authors evaluated the performance of the physical performance of the Short Physical Performance Battery (SPPB) in older adults living in 11 Peruvian high Andean communities. This work has a clear objective; the results are interesting and provide novel information in these regions of Peru with very particular characteristics given the geographical location and the height above sea level.\nI consider statistical analysis is proper to a cross sectional study using Poisson regression and was reported prevalence ratio with their confidence intervals.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "43044", "date": "12 Mar 2019", "name": "Tania Tello Rodriguez", "expertise": [ "Reviewer Expertise Frailty", "ageing" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIntroduction Older people from high Andean populations, in the majority of the cases, they do a lot of physical activity. However, the social context and the lack of health access services could be a negative factor for healthy ageing. It could be important to add something about physical activity in the introduction. A previous study of physical performance in older adults in rural areas in Peru found better physical performance in those who lived at height compared to those who lived at sea level, then probably height is not the cause to have a better or worse physical performance but other factors such as physical activity, multimorbidity, etc.\nMethods The best design for this kind of study is of case and control but the tranversal studies give us relevant information. The presence of osteoarthritis of knee and hip and low physical activity previous can impact in the physical performance in this study. In older people the BMI is not the best parameter to evaluate malnutrition and the recommended levels are different, as mentioned in this study. Describe the inclusion criteria and exclusion ones in detail, they are partially mentioned in the results.\nResults Related to the comorbility takes my attention the low percentage of arterial hypertension found in 10% and the high frequency of functional dependence is 83%. Very different amounts to that reported in previous studies.\nDiscussion 83% has functional dependence by the Barthel index, so it is a study to the community that is surprising, one of the hypothesis it is the population has high rates of multimorbility but to collect information but self-report the information was not given. There are other variables or conditions about the people living at high altitude that was missed in this study and that were not adjusted for in the regression models. The population that has cognitive impairment in a mild-moderate way by the test of Pfeiffer (screening test) could have a low score in the physical performance due to they did not understand in an appropriate way the orders to use SPPB; thus, in the context from people with low educational levels. The fact that they do not find links between poor physical performance and to live at high altitude support a previous study done1.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "44368", "date": "12 Mar 2019", "name": "Mariella Guerra", "expertise": [ "Reviewer Expertise Old age mental health" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nWas an interview for the informant to confirm participant data/information? Did you find participants who only speak Quechua or a dialect? If so did you translate the questionnaires? Differences between urban and rural areas? Need to discuss Peruvian research. Discussion must be done around community findings. You found no association between poor physical performance and disability. Explanations beside statistical results.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-59
https://f1000research.com/articles/8-1612/v1	09 Sep 19	{ "type": "Clinical Practice Article", "title": "A Hong Kong Chinese kindred with familial hypocalciuric hypercalcaemia caused by AP2S1 mutation", "authors": [ "Felix Chi Kin Wong", "Wai Sheung Wong", "Jeffrey Sung Shing Kwok", "Teresa Kam Chi Tsui", "Kam Piu Lau", "Michael Ho Ming Chan", "Yuet Ping Yuen", "Wai Sheung Wong", "Jeffrey Sung Shing Kwok", "Teresa Kam Chi Tsui", "Kam Piu Lau", "Michael Ho Ming Chan", "Yuet Ping Yuen" ], "abstract": "Familial hypocalciuric hypercalcaemia (FHH) is a genetic disorder of altered calcium homeostasis. Mutations in the CASR, GNA11 and AP2S1 genes have been reported to cause FHH. We report a Hong Kong Chinese kindred with FHH type 3 (FHH3) caused by mutations in AP2S1. The proband, a 51-year-old woman with hypercalcaemia, was initially diagnosed to have primary hyperparathyroidism but repeated parathyroidectomy failed to normalize her plasma calcium concentrations. Later, FHH was suspected and yet no mutations were identified in the CASR gene which causes FHH type 1 (FHH1), the most common form of FHH. Genetic testing of AP2S1 revealed a heterozygous c.43C>T (p.Arg15Cys) mutation, confirming the diagnosis of FHH3. The elder brother and niece of the proband, who both have hypercalcaemia, were found to harbour the same mutation. To our knowledge, this is the first Chinese kindred of FHH3 reported in the English literature.", "keywords": [ "Familial hypocalciuric hypercalcaemia type III", "AP2S1", "Hong Kong Chinese" ], "content": "Introduction\n\nFamilial hypocalciuric hypercalcaemia (FHH) is a genetically heterogeneous, autosomal dominant disorder characterized by a lifelong increase in plasma calcium concentrations with an inappropriately low urinary calcium excretion. In FHH, there is a reduction in the calcium-sensing ability of the chief cells of the parathyroid glands as well as an increase in tubular calcium reabsorption, resulting in an elevated homeostatic set-point of plasma calcium concentration and low urinary calcium excretion. Patients with FHH are generally asymptomatic, although some may develop pancreatitis or chondrocalcinosis1. Inactivating mutations in CASR was first reported in 1993 to cause FHH1, the most common form of FHH2. More recently, mutations in GNA113 and AP2S14 were identified to be responsible for FHH2 and FHH3, respectively. The following case report describes a Hong Kong Chinese kindred with hypercalcaemia and molecular diagnosis of FHH3. To our knowledge, this is the first Chinese kindred of FHH3 reported in the English literature.\n\n\nCase report\n\nThe proband was a 51-year-old Hong Kong Chinese woman who had an incidental finding of hypercalcaemia during hospitalization for Vibrio parahaemolyticus gastroenteritis. She presented with watery diarrhea, vomiting and colicky abdominal pain with dehydration. Results of laboratory tests on admission were as follows: plasma sodium 143 mmol/L (reference range [RR] 135 – 145 mmol/L), potassium 3.7 mmol/L (RR 3.5 – 5.1 mmol/L), creatinine 66 μmol/L (RR 44 – 80 μmol/L), urea 6.2 mmol/L (RR 2.7 – 6.8 mmol/L), total protein 83 g/L (RR 63 – 81 g/L) and albumin 49 g/L (RR 35 – 50 g/L). Unexpectedly, a grossly elevated plasma calcium level of 3.03 mmol/L (RR 2.10 – 2.55 mmol/L) was found. The plasma phosphate was 1.07 mmol/L (RR 0.90 – 1.55 mmol/L) and plasma alkaline phosphatase (ALP) was 97 U/L (RR 35 – 104 U/L). The patient had a past history of multi-nodular goiter and uterine fibroids with total hysterectomy and salpingo-oophorectomy performed in 1999. She did not have symptoms suggestive of hypercalcaemia, primary hyperparathyroidism or urolithiasis. She never took any calcium or vitamin supplements, and was not on any regular or over-the-counter medications.\n\nAfter rehydration, the plasma calcium was persistently elevated at 2.79 mmol/L. Whole blood ionized calcium was elevated at 1.47 mmol/L (RR 1.13 – 1.32 mmol/L), confirming genuine hypercalcaemia. The paired plasma parathyroid hormone (PTH) level was inappropriately normal (6.9 pmol/L, RR 1.7 – 9.2 pmol/L). Urinary calcium excretion over 24 hours was normal at 3.26 mmol/day (RR 2.50 – 8.00 mmol/day) with a urine volume of 2.23 L. Spot urine calcium and creatinine concentrations were also measured one day before the aforementioned 24-hour urine collection with paired measurement of plasma calcium and creatinine concentrations. The fractional excretion of calcium (FECa) was not documented in the patient’s medical record. In retrospect, it was low (0.55%, see Table 1, six months after presentation). Blood for serum protein electrophoresis, immunoglobulin pattern, erythrocyte sedimentation rate (ESR), and chest X-ray were unremarkable. Ultrasonography showed multi-nodular disease of thyroid gland and no parathyroid mass. Technetium-99m sestamibi parathyroid scintigraphy showed a small extra-thyroidal uptake focus near the lower pole of the right thyroid lobe suggestive of a right inferior parathyroid adenoma. A diagnosis of primary hyperparathyroidism due to right inferior parathyroid adenoma was made.\n\nBiochemical results measured on the same day are tabulated on the same row. The FECa of all affected family members was consistently below 1% (bold). *Age at which the patient first presented with hypercalcaemia. †Calculation of the fractional excretion of calcium (FECa) is as follows: ([urine calcium] × [plasma creatinine])/([urine creatinine] × [plasma calcium]) × 100%. ‡During the episode of acute pancreatitis. Reference ranges: plasma calcium (2.10 – 2.55 mmol/L), plasma phosphate (0.90 – 1.55 mmol/L), plasma creatinine (44 – 80 μmol/L), plasma PTH (1.7 – 9.2 pmol/L). Abbreviations: Ca, calcium; Cr, creatinine; FECa, fractional excretion of calcium; PO4, phosphate.\n\nRight inferior parathyroidectomy was performed 2 years after initial presentation. A 1 × 0.5 × 0.3 cm ovoid nodule was excised and frozen section showed a piece of parathyroid tissue without evidence of malignancy. However, the patient had persistent hypercalcaemia postoperatively with a plasma PTH level of 5.5 pmol/L. In addition, the surgical site was complicated by local haematoma formation. A second operation for haemostasis and subtotal parathyroidectomy was performed three days after the first operation. In order to maximize the probability of excising any hyperactive parathyroid tissue so that the normalization of plasma calcium level may be achieved, the right superior parathyroid gland and left inferior parathyroid gland were also excised, leaving only the left superior parathyroid gland in place. Histological examination confirmed further removal of parathyroid tissue with no evidence of parathyroid neoplasm. Nevertheless, her plasma calcium level remained elevated at levels of 2.63 to 2.84 mmol/L while the plasma PTH level remained non-suppressed at 4.6 pmol/L after the second operation. Technetium-99m sestamibi parathyroid scintigraphy three months after the second operation showed a suspicious right inferior hyperfunctioning parathyroid lesion, raising the possibility of residual parathyroid disease.\n\nAfter failing parathyroidectomy twice, the patient realized for the first time that her elder brother and her elder brother’s daughter also had incidental findings of hypercalcaemia (See Figure 1 for the pedigree and Table 1 for the summary of biochemical results). Upon further evaluation of the proband, 24-hour urine collection was repeated revealing a low urinary calcium excretion [2.09 mmol/day, urine volume 3.37 L, (RR 2.50 – 8.00 mmol/day)]. Spot urine calcium was 0.62 mmol/L and spot urine creatinine was 1.9 mmol/L. The concomitant plasma calcium was 2.63 mmol/l and plasma creatinine was 56 μmol/L. The FECa was 0.69%. A FECa of less than 1% is compatible with a diagnosis of FHH.\n\nThe proband is indicated by an arrow. The phenotype (plasma calcium concentration) as well as genotype of both parents of the proband is unknown. The plasma calcium concentration of the son of the proband [III(2)] was normal and targeted AP2S1 mutation analysis was negative. The calcium status of the baby boy of III(1) was not available and it was omitted from the pedigree.\n\nThe elder brother of the proband presented at the age of 54 years with an incidental finding of hypercalcaemia (plasma calcium: 2.80 mmol/L) during hospitalization for an episode of syncope. Measurement of 24-hour urinary calcium excretion was not performed. FECa was 0.83%. The niece of the proband presented at the age of 29 years with an incidental finding of hypercalcaemia (plasma calcium: 2.69 mmol/L) during hospitalization for urinary tract infection. Two years later, she had an episode of acute pancreatitis during pregnancy at 37 weeks’ gestation with serum amylase up to 1377 U/L. The concomitant triglyceride level was 1.7 mmol/L (RR <1.7 mmol/L) and plasma calcium was 2.65 – 2.69 mmol/L. Emergency Caesarean section was performed with uneventful delivery of a healthy baby boy, although the calcium status of her baby was unknown. Her pancreatitis responded to conservative management. Ultrasound of the neck showed no parathyroid nodules. Urinary calcium excretion over 24 hours was measured on two occasions but both results were within normal limits (4.92 and 3.18 mmol/day collected three years apart, with urine volumes of 3.3 and 3.5 L/day respectively). FECa was persistently low (0.09 – 0.75%).\n\nMutation analysis of the calcium-sensing receptor (CASR) gene was performed for the proband with all coding exons and flanking introns sequenced in both directions. No known pathogenic variants were identified. As a result, a diagnosis of FHH type 1 could not be confirmed. Sanger sequencing of the adaptor-related protein complex 2, sigma-1 subunit (AP2S1) gene showed heterozygous AP2S1 NM_004069.3: c.43C>T (p.Arg15Cys), which is a known pathogenic mutation of FHH3. The same AP2S1 mutation was identified in both her elder brother and niece, confirming the diagnosis of FHH3 in all three individuals. All three patients appeared to be cognitively normal. They were followed up for the monitoring of plasma calcium concentration. Up to one year after the genetic diagnosis was made, they remained asymptomatic and no treatment was given.\n\n\nDiscussion\n\nFHH is an important differential diagnosis of hypercalcaemia that one must carefully differentiate from primary hyperparathyroidism. The most indicative biochemical parameter for the diagnosis of FHH is the fractional excretion of calcium (FECa), also known as urinary calcium to creatinine clearance ratio. It is typically less than 1% in patients with FHH5–7. As exemplified in this case report, the diagnosis of FHH could be missed if 24-hour urinary calcium excretion only is taken into consideration. For calculation of the FECa, a spot urine sample or a 24-hour urine sample may be used, although 24-hour urine samples were originally employed to derive the cut offs5,6. Plasma PTH levels may be normal or raised in patients with FHH, similar to patients with primary hyperparathyroidism. Frequently, it remains difficult to distinguish primary hyperparathyroidism and FHH based on the available clinical, biochemical and radiological evidence, and the definitive diagnosis of FHH could only be achieved by genetic testing. Approximately 65% of individuals with FHH has FHH1, which is the most common type of FHH due to a loss-of-function of the calcium-sensing receptor (CaSR)8. More recently, mutations in GNA11 and AP2S1 have been identified to cause FHH2 and FHH3, respectively. FHH3 is caused by mutations at the Arg15 residue of the AP2S1 protein. Three different amino acid substitutions have been identified at this arginine residue, namely p.Arg15Cys (CGC➙TGC), p.Arg15Leu (CGC➙CTC) and p.Arg15His (CGC➙CAC). The Arg15 residue is highly evolutionarily conserved. In vitro functional studies suggest that a replacement of the positively charged Arg15 residue with the polar, but uncharged, Cys15 residue compromises the function of the adaptor-related protein complex 2 by reducing its affinity for the C-terminal calcium sensing receptor (CaSR) dileucine motifs, affecting the sensitivity of CaSR-expressing cells to extracellular calcium as well as resulting in the reduction of CaSR endocytosis4. In the same study, 11 individuals were found to harbour FHH3 mutations in 50 unrelated individuals with CASR mutation-negative FHH. Therefore, it is likely that AP2S1 mutations may be found in approximately 20% of patients of FHH without CASR mutations4. FHH2 is much rarer than FHH1 or FHH3. In fact, not a single case of FHH2 was detected in studies which collectively involved more than 200 patients with a phenotype of FHH9–11. Compared with patients with FHH1, patients with FHH3 are more likely to have higher serum calcium and magnesium, and lower fractional excretion of calcium. In addition, cognitive dysfunction, such as learning disability and attention deficit hyperactivity disorder, was seen in some patients with FHH3 and this was not seen in patients with FHH111,12. No apparent cognitive dysfunction was seen in any of our patients, and magnesium levels were not checked. Cinacalcet, a licensed calcium-sensing receptor allosteric activator, has been used successfully to reduce plasma calcium levels and hypercalcaemic symptoms in symptomatic FHH3 patients13. One of our patients (Figure 1, [III(1)]) developed an episode of acute pancreatitis during pregnancy, with no recurrence afterwards. Otherwise, they were free of hypercalcaemic complications or symptoms, and therefore no treatment was given.\n\nIn summary, we have reported the first Chinese kindred with FHH3 in the English literature. Mutation analysis of AP2S1 should be performed when the diagnosis of FHH is highly likely and yet no mutations in the CASR gene could be identified.\n\n\nConsent\n\nWritten informed consent for publication of their clinical details was obtained from all three patients.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.", "appendix": "References\n\nThakker RV: Diseases associated with the extracellular calcium-sensing receptor. Cell Calcium. 2004; 35(3): 275–282. PubMed Abstract \| Publisher Full Text\n\nPollak MR, Brown EM, Chou YH, et al.: Mutations in the human Ca2+-sensing receptor gene cause familial hypocalciuric hypercalcemia and neonatal severe hyperparathyroidism. Cell. 1993; 75(7): 1297–1303. PubMed Abstract \| Publisher Full Text\n\nNesbit MA, Hannan FM, Howles SA, et al.: Mutations affecting G-protein subunit α11 in hypercalcemia and hypocalcemia. N Engl J Med. 2013; 368(26): 2476–2486. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNesbit MA, Hannan FM, Howles SA, et al.: Mutations in AP2S1 cause familial hypocalciuric hypercalcemia type 3. Nat Genet. 2013; 45(1): 93–97. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMarx SJ, Attie MF, Levine MA, et al.: The hypocalciuric or benign variant of familial hypercalcemia: clinical and biochemical features in fifteen kindreds. Medicine (Baltimore). 1981; 60(6): 397–412. PubMed Abstract \| Publisher Full Text\n\nLaw WM Jr, Heath H 3rd: Familial benign hypercalcemia (hypocalciuric hypercalcemia). Clinical and pathogenetic studies in 21 families. Ann Intern Med. 1985; 102(4): 511–519. PubMed Abstract \| Publisher Full Text\n\nBrown EM: Clinical lessons from the calcium-sensing receptor. Nat Clin Pract Endocrinol Metab. 2007; 3(2): 122–133. PubMed Abstract \| Publisher Full Text\n\nNesbit MA, Hannan FM, Graham U, et al.: Identification of a second kindred with familial hypocalciuric hypercalcemia type 3 (FHH3) narrows localization to a <3.5 megabase pair region on chromosome 19q13.3. J Clin Endocrinol Metab. 2010; 95(4): 1947–1954. PubMed Abstract \| Publisher Full Text\n\nHovden S, Rejnmark L, Ladefoged SA, et al.: AP2S1 and GNA11 mutations - not a common cause of familial hypocalciuric hypercalcemia. Eur J Endocrinol. 2017; 176(2): 177–185. PubMed Abstract \| Publisher Full Text\n\nVargas-Poussou R, Mansour-Hendili L, Baron S, et al.: Familial Hypocalciuric Hypercalcemia Types 1 and 3 and Primary Hyperparathyroidism: Similarities and Differences. J Clin Endocrinol Metab. 2016; 101(5): 2185–2195. PubMed Abstract \| Publisher Full Text\n\nSzalat A, Shpitzen S, Tsur A, et al.: Stepwise CaSR, AP2S1, and GNA11 sequencing in patients with suspected familial hypocalciuric hypercalcemia. Endocrine. 2017; 55(3): 741–747. PubMed Abstract \| Publisher Full Text\n\nHannan FM, Howles SA, Rogers A, et al.: Adaptor protein-2 sigma subunit mutations causing familial hypocalciuric hypercalcaemia type 3 (FHH3) demonstrate genotype-phenotype correlations, codon bias and dominant-negative effects. Hum Mol Genet. 2015; 24(18): 5079–5092. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHowles SA, Hannan FM, Babinsky VN, et al.: Cinacalcet for Symptomatic Hypercalcemia Caused by AP2S1 Mutations. N Engl J Med. 2016; 374(14): 1396–1398. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "54581", "date": "16 Oct 2019", "name": "Caroline Gorvin", "expertise": [ "Reviewer Expertise Calcium-sensing receptor", "Hyper/Hypocalcaemia", "GPCRs", "Cell signalling" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nWong et al describe a Chinese kindred with FHH in which they have identified a previously described mutation in AP2S1. This is a comprehensive case report with detailed descriptions of the proband and several family members. Whilst not novel, this report describes FHH3 in a new ethnic population and further demonstrates that mutations in the Arg15 residue are the most commonly identified in FHH3. I have recommended some minor edits to improve the clarity of the report that are outlined below.\n\nAt present there are very few references in the introduction. Perhaps a few more could be inserted.\n\nPlease add an asterisk (or other appropriate symbol) next to statistically significant values in Table 1. This will make it immediately clear to the reader which values are significant.\n\nThe statement about FECa being <1% in FHH is made on page 3. This should be brought forward to page 2 following the first statement about the proband’s FeCa.\n\nThe second paragraph of page 2 begins with a statement about other members of the proband’s family, then goes on to discuss the proband again. I think this statement should be moved down to just before the mutational analysis. It will make the case report easier for a reader to understand if the proband is discussed first, then other family members later.\n\nIs the background of the cases’ history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the conclusion balanced and justified on the basis of the findings? Yes", "responses": [] }, { "id": "54582", "date": "21 Oct 2019", "name": "Fadil M. Hannan", "expertise": [ "Reviewer Expertise Genetics of calcium and parathyroid disorders", "molecular endocrinology of lactation" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe Clinical Practice Article by Wong and colleagues describes the first report of a Chinese kindred with familial hypocalciuric hypercalcaemia type 3 (FHH3). It is informative and well written, and provides detailed phenotypic information for this kindred. I have a few comments, which require addressing:\nTitle:\nPlease insert details of the mutation e.g. “…hypercalcaemia caused by an AP2S1 mutation, Arg15Cys”\n\nPage 3:\n\nIn the paragraph detailing parathyroid histology in the proband, is the term 'without evidence of malignancy', referring to parathyroid adenoma or carcinoma? Malignancy should not be used interchangeably with parathyroid adenoma, given that this is a benign condition.\n\nPlease avoid the term 'parathyroid neoplasm' as this suggests that histology is being undertaken to assess for parathyroid carcinoma.\n\nPlease cite Table 1 in the paragraph describing the findings in the elder brother.\n\nTable 1 should be cited after the statement 'FECa was 0.83%'.\n\nAlso cite Table 1 when describing the niece's biochemistry.\n\nPlease state if gallstones were excluded as a cause of the acute pancreatitis\n\nPage 4:\nThis article would benefit from a more in depth discussion of the pancreatitis in the niece. The authors should mention that pancreatitis has previously been reported in a child with FHH3 (Scheers et al Pancreatology 2019). This reported patient also harboured a mutation in the SPINK1 gene, which represents a risk factor for pancreatitis. The authors should indicate whether or not the niece has been tested for a mutation in the SPINK1 gene.\n\nTable 1:\nPlease insert serum 25-hydroxyvitamin D values for members of this kindred.\nFigure 1:\n\nPlease insert sequence trace for the proband showing the c.43C>T nucleotide substitution in the AP2S1 gene.\n\nIs the background of the cases’ history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Partly\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Partly\n\nIs the conclusion balanced and justified on the basis of the findings? Yes", "responses": [] }, { "id": "55718", "date": "01 Nov 2019", "name": "Pascal Houillier", "expertise": [ "Reviewer Expertise Disorders of mineral homeostasis" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nWong et al describe a Chinese kindred with FHH 3 due to a monoallelic point mutation in the AP2S1 gene. The phenotype of the proband and affected relatives is well described and the report is well written. I have a few comments that I submit to the consideration of the authors:\nThe work-up of patients with primary hyperparathyroidism commonly includes renal ultrasonography (or CT scan) and bone mineral density measurement. Was any of those performed in any of the patients? If so, the results should be provided. If not, this should be mentioned.\n\nA score, named Pro-FHH,1 has recently been reported as performing better that CCCR to distinguish patients with primary hyperparathyroidism and with FHH. Could Pro-FHH be retrospectively be computed in the patients? The respective merits of CCCR and Pro-FHH could be discussed.\n\nThe description of the parathyroid tissue is a bit confusing. Is neoplasm used instead of carcinoma? Was the removed parathyroid tissue totally normal (on a quantitative and qualitative basis), or hyperplastic or anything?\n\nTable 1 should be cited when describing the biochemistry in the brother and the niece.\n\nHospitalization for urinary tract infection is uncommon unless it is complicated. Was it?\n\nThe risk of post surgical hypoparathyroidism after repeated parathyroid surgery should be mentioned in the discussion.\n\npage 1, introduction: renal tubular calcium reabsorption\n\nIs the background of the cases’ history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Partly\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Partly\n\nIs the conclusion balanced and justified on the basis of the findings? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1612
https://f1000research.com/articles/7-1908/v1	07 Dec 18	{ "type": "Data Note", "title": "Pan-cancer repository of validated natural and cryptic mRNA splicing mutations", "authors": [ "Ben C. Shirley", "Eliseos J. Mucaki", "Peter K. Rogan", "Ben C. Shirley", "Eliseos J. Mucaki" ], "abstract": "We present a major public resource of mRNA splicing mutations validated according to multiple lines of evidence of abnormal gene expression. Likely mutations present in all tumor types reported in the Cancer Genome Atlas (TCGA) were identified based on the comparative strengths of splice sites in tumor versus normal genomes, and then validated by respectively comparing counts of splice junction spanning and abundance of transcript reads in RNA-Seq data from matched tissues and tumors lacking these mutations. The comprehensive resource features 351,423 of these validated mutations, the majority of which (69.1%) are not present in the Single Nucleotide Polymorphism Database (dbSNP 150). There are 117,951 unique mutations which weaken or abolish natural splice sites, and 244,415 mutations which strengthen cryptic splice sites (10,943 affect both simultaneously). 27,803 novel or rare flagged variants (with <1% population frequency in dbSNP) were observed in multiple tumor tissue types. Single variants or chromosome ranges can be queried using a Global Alliance for Genomics and Health (GA4GH)-compliant, web-based Beacon “Validated Splicing Mutations” either separately or in aggregate alongside other Beacons through the public Beacon Network (http://www.beacon-network.org/#/search?beacon=cytognomix), as well as through our website (https://validsplicemut.cytognomix.com/).", "keywords": [ "RNA Splice Sites", "Single Nucleotide Polymorphism", "Genome", "Mutation", "Chromosomes", "Neoplasms", "Information Theory", "Next generation sequencing", "validation" ], "content": "Introduction\n\nNext generation sequencing continues to reveal large numbers of novel variants whose impact cannot be interpreted from curated variant databases, or through reviews of peer-reviewed biomedical literature1. This has created a largely unmet need for unequivocal sources of information regarding the molecular phenotypes and potential pathology of variants of unknown significance (VUS); in cancer genomes, such sources are critically needed to assist in distinguishing driver mutations from overwhelming numbers of bystander mutations. VUS classification criteria highlight the limitations in genome interpretation due to ambiguous variant interpretation. Of the 458,899 variant submissions in NCBI’s ClinVar database with clinical interpretations, nearly half (n=221,271) are VUS (as of November 5th 2018). Only 10,784 variants in ClinVar have been documented to affect mRNA splicing at splice donor or acceptor sites, with 1,063 of these being classified as VUS, and cryptic mRNA splicing mutations are not explicitly described. The current ACMG criteria2 for variant pathogenicity prevent clinical classification of most VUS. Functional evidence that VUS either disrupt or abolish expression of genes has been sought to improve classification and provide insight into the roles, if any, of individual VUS in predisposing or causing disease. We present a comprehensive data repository for a relatively common mutation type (cis-acting variants that alter mRNA splicing). Mutations are predicted with information theory-based analyses3, and supported with functional evidence that variants in tumor genomes are specifically associated with abnormally spliced mRNAs that are infrequent or absent in transciptomes lacking these variants4.\n\nInformation theory (IT) has been proven to accurately predict impact of mutations on mRNA splicing, and has been used to interpret coding and non-coding mutations that alter mRNA splicing in both common and rare diseases3,5–15. We have described an IT-based framework for the interpretation and prioritization of non-coding variants of uncertain significance, which has been validated in multiple studies involving novel variants in patients with history or predisposition to heritable breast and/or ovarian cancer11–15.\n\nThe Cancer Genome Atlas (TCGA) Pan-Cancer Atlas (PCA) is a comprehensive integrated genomic and transcriptomic resource containing data from >10,000 tumors across 33 different tumor types16. Here, we utilized IT-based tools for assessment of high quality sequenced variants in TCGA patients for their potential impact on mRNA splicing. The accuracy of predicted mutations was evaluated with an algorithm we previously developed that compares transcripts from individuals carrying these variants with others lacking them. The results of these genome-wide analyses are presented using an online resource which can be queried through the Beacon Network17.\n\n\nMethods\n\nControlled-access data was obtained with permission from the Data Access Committee at NIH for TCGA and from the International Cancer Genomics Consortium. Patient RNA sequencing BAM files (tumor and normal, when available) and their associated VCF files (GRCh37) were initially obtained from the CancerGenomeHub (CGhub). Files were later downloaded through Genomic Data Commons using the GDC Data Transfer Tool (version 1.3.0), as CGhub was decommissioned mid-project. Variants in VCF files which did not pass quality control (QC) were not analyzed.\n\nWe used the Shannon Pipeline software (which applies IT to rapidly perform high-throughput, in silico prediction of the impacts of variants on mRNA splicing)18 to analyze all QC-passing variants in VCFs from TCGA (>168 million variants) to evaluate their potential impact on splice site binding strength (changes in information content, Ri, measured in bits). Variants which were predicted to strengthen known natural sites or weaken cryptic splice sites were excluded from all subsequent analyses.\n\nTo validate the potential impact of Shannon Pipeline-flagged mutations, Veridical software analyzed genomic variants (including insertions and deletions) by comparing the RNA-Seq alignment in the region surrounding the variant with the corresponding interval in control transcriptomes (normal and tumor tissue of the same type) lacking the variant4,19. Veridical: a) counts abnormally spliced reads in RNA-Seq data (categorized as: cryptic site use, exon skipping, or intron inclusion [containing or adjacent to the flagged mutation]), b) applies the Yeo-Johnson transformation to these results, and c) determines the null hypothesis probability (p-value) that the transformed read count corresponds to normal splicing. In tumor types where normal controls were not available, a set of RNA-Seq datasets from 100 different normal tissues from TCGA were used (e.g. a combination of 5 tissue types: BRCA, BLCA, LUAD, KIRC, PRAD). Veridical results that were not significant for a particular variant (p-value > 0.05 for all of the splicing categories) were not further analyzed. After analysis, Veridical validated 351,423 unique mutations for their direct impact on mRNA splicing (Table 1). The Shannon pipeline-flagged and Veridical-filtered results were combined into a single large table (Dataset 120), the source data for the ValidSpliceMut SQL database and the associated Beacon application.\n\n*The number of Veridical-flagged mutations in each The Cancer Genome Atlas (TCGA) cancer data set. Variants shared between multiple tissue types are counted for each category. Variant and RNA-Seq data were provided by The Cancer Genome Atlas Pan-Cancer Analysis Project16.\n\nWe created a publicly accessible Application Programming Interface (API) (https://beacon.cytognomix.com) that can be utilized to programmatically query variants passing filter thresholds described above (Dataset 120). It was built in accordance with the GA4GH Beacon v1.0.0 specification, which describes a Representational State Transfer (REST) API for genetic data sharing. A Beacon accepts queries using an HTTP request and returns JavaScript Object Notation (JSON). Our Beacon implementation is coded in PHP 7.0 and utilizes a MySQL database (version: 5.7.24) with indexes applied to variant ID, chromosome, and coordinate fields (GRCh37). The returned JSON object reports whether the variant was found within our Beacon dataset as well as metadata including splice site coordinate, splice type, site type, the IT-based measures Ri,initial and Ri,final, affected individual IDs, tumor type, Veridical evidence by type annotated with significance level, and, if known, the corresponding rsID with its average heterozygosity (dbSNP 150). The metadata for each variant sent to the Beacon Network is a concise subset of available results in our database. It includes the first relevant database entry, meaning that if the variant exists within multiple individuals only the first will contribute fields to the metadata. However, among this metadata is a hyperlink to our local website containing results for any remaining tumors.\n\nWe developed the website ValidSpliceMut (Figure 1) to serve as a local interface to our Beacon, allowing users to manually search for a variant, by gene name or genome coordinate range. ValidSpliceMut automatically queries our Beacon, and formats the results of the search, if any. This website provides a complete view of variants, including Veridical-based evidence on all data related to every affected individual. If a variant is associated with multiple splice sites, the user is presented with a brief overview of all affected sites and must select a desired site to continue. To obtain the coordinate of the queried variant in gene-centric notation, a link is provided which queries the Mutalyzer API and generates coordinates for all available transcripts. ValidSpliceMut only reports transcripts for the gene affected by the variant.\n\n(A) The ‘Variant Position’ heading displays the variant of interest in g. notation, and provides a link which queries the Mutalyzer API to obtain the variant coordinate in a gene-centric c. mutation format. Variant-specific and splice site-specific tabular results are presented under the headings “Splice Site Information” and “Variant Data”. Results are organized by TCGA sample IDs harboring the mutation within a series of expandable panels. A link is provided to patient tumor metadata on the GDC data portal. Each panel consists of read counts and p-values by Veridical evidence type. Significant p-values (≤ 0.05) are highlighted in bold. Evidence types deemed “strongly corroborating” (Viner et al. 2014) are color coded and correspond to the dynamically generated text appearing above the table. (B) An integrative genome viewer (IGV) image showing alignment of expressed sequence reads. IGV screenshots are provided only for mutations present <1% of population (in dbSNP 150), with ≥ 5 junction-spanning reads, and are highly significant (p < 0.01) for cryptic splicing, exon skipping, and/or intron inclusion with mutation. A specific IGV screenshot for this sample captures the region surrounding the mutation. Here, several RNA-Seq reads show skipping of the affected exon. (C) A dynamically generated histogram presents expression levels of all genes for a selected normal tissue type. Genes are grouped into bins based on expression level, denoted on the x-axis. The number of genes present in each bin is shown on the y-axis (log10 scale). The histogram key indicates the expression range of the variant-containing gene. Tissue type can be changed via a drop-down list.\n\nA results page presents variant-specific data in tabular format and an expandable list of panels describing the affected individuals. Each of these panels contains Veridical output in tabular format for the selected tumor, a link to the tumor metadata at US National Cancer Institute (by querying the GDC API to obtain a UUID which is used to construct a link to the GDC data portal), an Integrative Genome Viewer (IGV) screenshot containing the variant (IGV screenshots are available for selected variants, see below), and a histogram which presents the expression levels of the variant-containing gene compared to all other gene expression levels across a selected normal tissue type (created dynamically using gnuplot 5.0). The tissue expression data is provided by GTEx (downloaded on 10/22/18). However, several TCGA tumor types did not have a GTEx equivalent (CHOL, DLBC, MESO, READ, SARC, THYM and UVM). The GNF Expression Atlas 221 was downloaded from the UCSC Genome Browser and was used for expression data for both lymph nodes (DLBC) and the thymus (THYM). For the remaining tissues, expression data from the following studies were obtained from the Genome Expression Omnibus (GEO): GSE76297 (CHOL), GSE2549 (MESO), GSE15781 (READ) GSE44426 (SARC), and GSE44295 (UVM).\n\nTo generate IGV images presented on the webpage, a bash script was written to automatically load the RNA-Seq BAM file of a patient with a mutation of interest into IGV, set the viewing window within the region of interest (300nt window, centered on the variant), sorted to bring reads containing the variant of interest to the top of the screen (to increase chance of visualizing mutant splice form), followed by a screen capture. The generation and storage of IGV images for all patient-mutation pairs would be prohibitive due to limitations in time and server space requirements. Therefore IGV images showing evidence of splicing abnormalities were generated only for patient-mutation pairs which met the most stringent criteria: the mutation was required to be flagged for junction-spanning cryptic site use, exon skipping, or intron inclusion (with mutation); the flagged category must include 5 or more reads in this category; if the variant is present in the dbSNP database (release 150), the frequency was required to be < 1% of the population; and the Veridical results, in which the mutations flagged were required to exhibit p ≤ 0.01 for at least one form of evidence of a splicing abnormality. In some cases, the splicing event observed by Veridical may not be present within the image window as the automated procedure used to create these images does not present all evidential sequence reads due to limitations on the number of reads that are shown. Additionally, reads appearing as exon skipping may instead indicate a pre-existing cryptic site outside of the viewing window (see Table 2; FAT1:g.187521515C>A [c.11641-1G>T] and SMAD3:g.67482748C>G [c.1155-3C>G]).\n\nExample mutations which alter splicing in tumor-associated genes found in patients with the same tumor type. Mutations are linked to their page on https://validsplicemut.cytognomix.com/, which provides additional material such as RNAseq images of the regions of interest. GRCh37 coordinates provided.\n\n\nDataset validation and discussion\n\nWe have derived a GA4GH-standardized, searchable web resource for a large set of validated mRNA splicing variants present in diverse tumor types. All variants passing QC in TCGA cancer patients were analyzed with the Shannon pipeline18. This revealed that 1,297,242 variants were predicted to have significant impacts on normal mRNA splicing (347,549 natural and 985,112 cryptic splice sites; 35,419 affecting both types). Subsequent RNA-Seq analysis with Veridical4 provided evidence of abnormal gene expression specifically associated with a subset of these variant(s), identifying 351,423 unique mutations. Results are searchable through either the Beacon Network, or our publicly-accessible webpage.\n\nOur results contrast with another TCGA study that investigated alternative mRNA splicing22 and demonstrated a limited set of non-constitutive exon-exon junctions attributable to cis-acting splicing mutations (n = 32). The 2,736 novel or rare variants that we report which specifically activate cryptic splicing (significant ‘junction-spanning cryptic site use’ reads found by Veridical), exceed the number reported in another study that analyzed all available TCGA tumor transcriptomes (n=1,964)23.\n\nValidated variants (which we define as mutations) were also tallied by tumor tissue type in our study (Table 1). 33.6% of unique mutations (n=117,951) significantly weaken natural splice sites, while 69.6% (n=244,415) strengthen novel or pre-existing cryptic sites. 242,983 mutations (69%) are absent from dbSNP 150. 73,975 mutations (21%) are present in <1% of the population, of which 27,803 of these (and those not present in dbSNP) were present in multiple tumor types. Valid mutations lacking rsIDs represent either novel or recently observed variants. This low level of dbSNP saturation is consistent with the idea that many currently unknown mRNA splicing mutations may yet be discovered through additional sequencing studies.\n\nIn Table 2, we highlight a subset of validated splicing mutations (n=25) which were identified in known driver genes implicated in the COSMIC (Catalogue Of Somatic Mutations In Cancer) Cancer Gene Census catalog (CGC)24. These mutations are associated with either increased exon skipping, intron inclusion, and/or cryptic site use. Mutations in Table 2 are hyperlinked to the ValidSpliceMut webpage which provides additional information, including expression evidence supporting predictions made by the Shannon pipeline.\n\nMany mutations generated multiple types of abnormal read evidence present in mis-spliced transcripts. Interestingly, a subset of mutations (n=28) produced evidence for every type of abnormal splicing reported by Veridical. Dataset 225 (see Data Availability) describes 11 representative mutations that simultaneously increase exon skipping, intron inclusion, and activate (or significantly increase utilization of) a strengthened cryptic site. In all but one instance, the mutation weakens the natural site while simultaneously strengthening a nearby cryptic site. The one exception involves the gene SAP30BP, where simultaneously occurring mutations in the same read (in linkage disequilibrium; separated by 4 nucleotides) independently cause two separate splicing changes: g.73702087G>A (c.661-1G>A; abolishes the natural acceptor of exon 10) and g.73702091G>A (c.664G>A; creates a weak cryptic acceptor site). The combined splicing impact of these variants is significant exon skipping, intron inclusion, and use of the activated cryptic site.\n\nBecause of the requirement for expression validation, this resource presents a set of splicing abnormalities in which we have the highest confidence. We anticipate that some correct predictions of the Shannon pipeline may have not been validated by Veridical due to the limitations of mRNA detection; for example, either low expression of the gene harboring the mutation or nonsense-mediated decay of the corresponding transcript could be consistent with the effects of a valid splicing mutation, but in the absence of a sufficient number of abnormal reads, the mutation could not be confirmed. Furthermore, at the time that the current analysis was performed, the available Shannon pipeline version did not report regulatory splicing variants adjacent to constitutive and cryptic splice sites which influence exon definition. Due to the substantial processing required for the complete TCGA dataset, the present analysis does not incorporate the effects of these variants on exon definition, which we have modeled by IT6; it does not predict the relative abundance of leaky, natural and cryptic isoforms, though such information might be inferred from the expression data on each tumor. The current version of Shannon pipeline does integrate predictions of splicing regulatory sequences and accounts for relative abundance of mRNA isoforms by exon definition, and is available through the MutationForecaster system.\n\nThe Validated Splicing Mutation resource should substantially contribute to reducing the number of outstanding VUS in tumor (and possibly some germline) genomes, and substantially increases the number of functional variants with previously unappreciated consequences to mRNA splicing, in particular, activation of cryptic splice sites. In our previous study19, a subset of the TCGA breast cancer patient data was evaluated with IT-based tools, identifying 988 mutations as significantly altering normal splicing by Veridical (19% of total mutations flagged by IT). This database greatly expands the size of the repository. Here, a higher ratio of rare or novel mutations have been validated by Veridical (24% of total mutations were flagged by IT). The higher yield found could be related to the same mutation being present in multiple samples from the same tumor type and other tumor tissues, which would be expected to increase the probability of observing abnormally expressed splice forms for the mutation.\n\n\nbioRxiv\n\nAn earlier version this article is available from bioRxiv: https://doi.org/10.1101/47445226\n\n\nSoftware availability\n\nArchieved code and scripts used as part of this study are available from Zenodo,\n\nZenodo: Validated Splicing Mutations Beacon API http://doi.org/10.5281/zenodo.157989827\n\nZenodo: Validated Splicing Mutations Website http://doi.org/10.5281/zenodo.157982228\n\nZenodo: Expression Data Processing, Histogram input generation and IGV Bash Script Generating Programs http://doi.org/10.5281/zenodo.158242129\n\nAll software is licensed under a Creative Commons Attribution-Non Commercial-ShareAlike 4.0 International Public License\n\n\nData availability\n\nZenodo: Dataset 1. Validated natural and cryptic mRNA splicing mutations. Source data computed by the Shannon pipeline and Veridical, displayed on the ValidSpliceMut website (https://validsplicemut.cytognomix.com/). DOI: http://doi.org/10.5281/zenodo.148821120\n\nZenodo: Dataset 2. Mutations which lead to multiple types of aberrant splicing. Representative set of mutations which significantly alter splicing in all evidence types analyzed by Veridical (i.e. cryptic splice site use, exon skipping, intron inclusion). Mutations are linked to their page on https://validsplicemut.cytognomix.com/, which provides additional material such as RNA-Seq images of the regions of interest. DOI: https://dx.doi.org/10.5281/zenodo.148994125\n\nLicense: CC0 1.0\n\n\nConsent\n\nControlled-access TCGA sequence data was accessed with permission from NCBI (dbGaP Project #988: “Predicting common genetic variants that alter the splicing of human gene transcripts”; Approval Number #13930-11; PI: PK Rogan) and the International Cancer Genome Consortium (ICGC Project #DACO-1056047; “Validation of mutations that alter gene expression”).", "appendix": "Grant information\n\nPKR is supported by The Natural Sciences and Engineering Research Council of Canada (NSERC) [RGPIN-2015-06290], Canadian Foundation for Innovation, Canada Research Chairs, and CytoGnomix. Compute Canada and Shared Hierarchical Academic Research Computing Network (SHARCNET) provided high performance computing and storage facilities.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe acknowledge Coby Viner, Stephanie Dorman, William J. Phillips and Ujani Hazra for their contributions to the early stages of this project. We are grateful to Max Barkley, Milan Panik and Miro Cupak (DNAStack) for their assistance in integrating our ValidSpliceMut Beacon into the GA4GH network.\n\n\nReferences\n\nFoley SB, Rios JJ, Mgbemena VE, et al.: Use of Whole Genome Sequencing for Diagnosis and Discovery in the Cancer Genetics Clinic. EBioMedicine. 2015; 2(1): 74–81. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRichards S, Aziz N, Bale S, et al.: Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015; 17(5): 405–424. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCaminsky N, Mucaki EJ, Rogan PK: Interpretation of mRNA splicing mutations in genetic disease: review of the literature and guidelines for information-theoretical analysis [version 1; referees: 2 approved]. F1000Res. 2014; 3: 282. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nViner C, Dorman SN, Shirley BC, et al.: Validation of predicted mRNA splicing mutations using high-throughput transcriptome data [version 2; referees: 4 approved]. F1000Res. 2014; 3: 8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMucaki EJ, Ainsworth P, Rogan PK: Comprehensive prediction of mRNA splicing effects of BRCA1 and BRCA2 variants. Hum Mutat. 2011; 32(7): 735–742. PubMed Abstract \| Publisher Full Text\n\nMucaki EJ, Shirley BC, Rogan PK: Prediction of mutant mRNA splice isoforms by information theory-based exon definition. Hum Mutat. 2013; 34(4): 557–565. PubMed Abstract \| Publisher Full Text\n\nRogan PK, Svojanovsky S, Leeder JS: Information theory-based analysis of CYP2C19, CYP2D6 and CYP3A5 splicing mutations. Pharmacogenetics. 2003; 13(4): 207–218. PubMed Abstract\n\nRogan PK, Schneider TD: Using information content and base frequencies to distinguish mutations from genetic polymorphisms in splice junction recognition sites. Hum Mutat. 1995; 6(1): 74–76. PubMed Abstract \| Publisher Full Text\n\nRogan PK, Faux BM, Schneider TD: Information analysis of human splice site mutations. Hum Mutat. 1998; 12(3): 153–171. PubMed Abstract \| Publisher Full Text\n\nPeterlongo P, Catucci I, Colombo M, et al.: FANCM c.5791C>T nonsense mutation (rs144567652) induces exon skipping, affects DNA repair activity and is a familial breast cancer risk factor. Hum Mol Genet. 2015; 24(18): 5345–5355. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMucaki EJ, Caminsky NG, Perri AM, et al.: A unified analytic framework for prioritization of non-coding variants of uncertain significance in heritable breast and ovarian cancer. BMC Med Genomics. 2016; 9: 19. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCaminsky NG, Mucaki EJ, Perri AM, et al.: Prioritizing Variants in Complete Hereditary Breast and Ovarian Cancer Genes in Patients Lacking Known BRCA Mutations. Hum Mutat. 2016; 37(7): 640–652. PubMed Abstract \| Publisher Full Text\n\nYang XR, Devi BCR, Sung H, et al.: Prevalence and spectrum of germline rare variants in BRCA1/2 and PALB2 among breast cancer cases in Sarawak, Malaysia. Breast Cancer Res Treat. 2017; 165(3): 687–697. PubMed Abstract \| Publisher Full Text\n\nDos Santos ES, Caputo SM, Castera L, et al.: Assessment of the functional impact of germline BRCA1/2 variants located in non-coding regions in families with breast and/or ovarian cancer predisposition. Breast Cancer Res Treat. 2018; 168(2): 311–325. PubMed Abstract \| Publisher Full Text\n\nBurke LJ, Sevcik J, Gambino G, et al.: BRCA1 and BRCA2 5’ noncoding region variants identified in breast cancer patients alter promoter activity and protein binding. Hum Mutat. 2018; 39(12): 2025–2039. PubMed Abstract \| Publisher Full Text\n\nHoadley KA, Yau C, Hinoue T, et al.: Cell-of-Origin Patterns Dominate the Molecular Classification of 10,000 Tumors from 33 Types of Cancer. Cell. 2018; 173(2): 291–304.e6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGlobal Alliance for Genomics and Health: GENOMICS. A federated ecosystem for sharing genomic, clinical data. Science. 2016; 352(6291): 1278–1280. PubMed Abstract \| Publisher Full Text\n\nShirley BC, Mucaki EJ, Whitehead T, et al.: Interpretation, stratification and evidence for sequence variants affecting mRNA splicing in complete human genome sequences. Genomics Proteomics Bioinformatics. 2013; 11(2): 77–85. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDorman SN, Viner C, Rogan PK: Splicing mutation analysis reveals previously unrecognized pathways in lymph node-invasive breast cancer. Sci Rep. 2014; 4: 7063. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMucaki EJ, Shirley BC, Rogan PK: Dataset 1. Validated natural and cryptic mRNA splicing mutations [Data set]. Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1488211\n\nSu AI, Wiltshire T, Batalov S, et al.: A gene atlas of the mouse and human protein-encoding transcriptomes. Proc Natl Acad Sci U S A. 2004; 101(16): 6062–6067. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKahles A, Lehmann KV, Toussaint NC, et al.: Comprehensive Analysis of Alternative Splicing Across Tumors from 8,705 Patients. Cancer Cell. 2018; 34(2): 211–224.e6. PubMed Abstract \| Publisher Full Text\n\nJayasinghe RG, Cao S, Gao Q, et al.: Systematic Analysis of Splice-Site-Creating Mutations in Cancer. Cell Rep. 2018; 23(1): 270–281.e3. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFutreal PA, Coin L, Marshall M, et al.: A census of human cancer genes. Nat Rev Cancer. 2004; 4(3): 177–183. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMucaki EJ, Shirley BC, Rogan PK: Dataset 2. Mutations which lead to multiple types of aberrant splicing. Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1489941\n\nShirley BC, Mucaki EJ, Rogan PK: Pan-Cancer Repository of Validated Natural and Cryptic mRNA Splicing Mutations. bioRxiv. 2018; 474452. Publisher Full Text\n\nShirley BC, Mucaki EJ, Rogan PK: Validated Splicing Mutations Beacon API (Version 1.0.0). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1579898\n\nShirley BC, Mucaki EJ, Rogan PK: Validated Splicing Mutations Website (Version 1.0.0). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1579822\n\nMucaki EJ, Shirley BC, Rogan PK: Expression Data Processing, Histogram input generation and IGV Bash Script Generating Programs. Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1582421" }	[ { "id": "41666", "date": "12 Dec 2018", "name": "Emanuele Buratti", "expertise": [ "Reviewer Expertise I have more than twenty years experience in the investigation of pre-mRNA splicing processes and especially their potential connection with a variety of human diseases", "both monogenic (Cystic Fibrosis", "Pompe Disease", "Neurofibromatosis) and polygenic (Amyotrophic Lateral Sclerosis", "Frontotemporal Dementia). I am the author of more than 160 research papers in peer-reviewed publications and of several articles in scientific books on these subjects (orcid.org/0000-0002-1356-9074)" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe increasing amount of sequencing data that is being generated in many biological systems has represented a real challenge to researchers in terms of trying to link individual changes to a particular biological process. The attempt, described in this work, to use IT approaches to evaluate the potential biological significance can significantly contribute to fill this gap. The laboratory of Peter Rogan has a long standing and internationally prominent role in addressing the possible consequences of sequence variants on the pre-mRNA splicing process especially with regards to its connection with human disease. The ValidSpliceMut developed in this work presents a user friendly interface that allows users to manually search for a variant (by gene name or genome coordinate range) and obtain information with regards to its possible effect on splicing. This will greatly help to better appreciate the functionality of Variants of Unknown Significance that are currently abundant genomic and transcriptomic Atlases.\n\nIs the rationale for creating the dataset(s) clearly described? Yes\n\nAre the protocols appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and materials provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Yes", "responses": [] }, { "id": "41665", "date": "07 Jan 2019", "name": "Francesca D. Ciccarelli", "expertise": [ "Reviewer Expertise Computational cancer genomics" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper entitled \"Pan-cancer repository of validated natural and cryptic mRNA splicing mutations\" by Shirley, Mucaki and Rogan describes an extensive analysis of pancancer somatic variants in samples of the TCGA dataset to identify mutations that affect splicing. To this aim, the authors combine the methods that they previously developed to predict mutations affecting splicing, with a validation of their effect on matched mRNA data from TCGA.\n\nThe study is technically sound and follows a line of investigation that has been a long standing interest of the authors. Despite this, I have a number of comments that hopefully will help strengthen the study:\nThe authors write that their IT-based framework to predict slicing variants \"has been validated in multiple studies\" and they refer to numerous papers. However, in all of them they act as co-authors, showing that their method is mostly used by themselves and their collaborators. This is not necessarily a problem, but it would certainly strengthen the study if the authors would perform a comparative assessment of their performance with other available methods to predict splicing mutations, for example those in dbNSFP. This will provide a less biased interpretation of the final results. Somehow related to the previous point, the authors mention that their results \"contrast with another TCGA study that investigated alternative mRNA splicing\". In my opinion this point should be further explored: what are the main differences and what is the extent of overlap in concordant predictions? What are the possible reasons for these differences? This is important because the cited paper in Cancer Cell analysed the same dataset of mutations. The authors notice that the number of variants which activate cryptic splicing exceed the number reported in a recently published study in Cell Reports. Similarly to before: what is the extent of overlap between the two datasets? Stating that a dataset is bigger than another one is not necessarily an indication that it is better. The authors validate ~27% of predicted splicing variants using the mRNA data (351k validated of the 1.2M predicted). This is a surprisingly low fraction. Later in the manuscript, the authors briefly discuss about the possible reasons of such a discrepancy. One of them is the possible occurrence of nonsense mediate decay which will not confirm the mutations because no or very few reads will be detected. However, as the authors acknowledge, the absence of supporting reads only in mutated individuals as compared to the presence of reads in WT sample would be a strong indication of the effective role of these mutations on splicing. This can be quantified from the same RNAseq data and in my opinion should be done. In general, the authors seem to exclude that their prediction method could lead to false positives. Rather they justify the poor overlap with limitations of mRNA detection. If this is the case, this should be quantified and probably a comparison with other prediction methods could help. Of the >351k mutations with an effect on splicing supported by RNA data, only 35 affect CGC genes. Is this only a subset of mutations affecting driver genes or is it the complete list? In the former case, I would suggest that the authors provide the full list as supplementary data. In the latter case, the authors should discuss the implication of such a low number. Considering that there are >700 CGC genes, does it mean that aberrant slicing is very rarely a driver event? Is the overwhelming majority of splicing variants passenger?\n\nIs the rationale for creating the dataset(s) clearly described? Yes\n\nAre the protocols appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and materials provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Yes", "responses": [ { "c_id": "4469", "date": "20 Mar 2019", "name": "Peter Rogan", "role": "Author Response", "response": "We thank the reviewer for their valuable comments. Our responses follow: 1. The authors write that their IT-based framework to predict slicing variants \"has been validated in multiple studies\" and they refer to numerous papers. However, in all of them they act as co-authors, showing that their method is mostly used by themselves and their collaborators. This is not necessarily a problem, but it would certainly strengthen the study if the authors would perform a comparative assessment of their performance with other available methods to predict splicing mutations, for example those in dbNSFP. This will provide a less biased interpretation of the final results. Response: We have previously compared our mutation prediction methods with others. Mucaki et al. Hum. Mut. 34:556-65 (2013) showed that IT-based single mutation and exon definition methods performed as well or better than MaxEntScan and Human Splice Finder websites. MaxEntScan computes relative entropy which is similar to IT, except it applies a correction for local base composition (which does not measure free energy, in contrast with IT: J. Theor. Biol. 201:87-92, 1989). Human Splice Finder does not measure changes in binding affinity and its basis is ad hoc. We also reviewed all articles (300+) which have used IT-based tools to predict changes in splicing (Caminsky et al., F1000Research 3:282, 2014). This reference covers the vast majority of studies that used IT-based bioinformatic tools for mutation analysis and, compared results obtained using these tools with other available software. The cited studies included a large proportion of mutations that we, ourselves, did not coauthor, or were analyzed by others, removing an obvious source of bias. Regarding potential bias in our results, the IT-based position weight matrices (iPWMs) of splice recognition sites that we derived and use are based on a comprehensive set of splice sites spanning all known coding genes (see appendix of Rogan et al. Pharmacogenetics & Genomics 13(4):207-18, 2003). Other bioinformatic methods for splice site detection are based on many fewer splice sites for PWMs and are much more likely to be subject to bias based on how those sites were chosen. Also, the determination of information content in natural and mutated splice sites obeys the second law of thermodynamics (Schneider, J. Theor. Biol. 189:427-41, 1997); information contents have been formally proven to be related to binding affinities of splice site to splicesomes and splicing factors. MaxEntScan differs from IT because it applies a correction for local base composition, which is energetically and biochemically irrelevant to binding site affinity, and is therefore, biased. As comparisons of IT with other methods have been covered previously, and the F1000Research Data Note article format is intended to specifically present results using the new resource we describe, in our opinion, reevaluation of other algorithms would not add anything of value to this manuscript. 2. Somehow related to the previous point, the authors mention that their results \"contrast with another TCGA study that investigated alternative mRNA splicing\". In my opinion this point should be further explored: what are the main differences and what is the extent of overlap in concordant predictions? What are the possible reasons for these differences? This is important because the cited paper in Cancer Cell analysed the same dataset of mutations. Response: The paper being referred to in the reviewer’s comment is Kahles et al., Cancer Cell., 34(2):211–224.e62018. This paper reports: “In a joint analysis of cis and trans associations with 50% prior on each type, we identified 32 cis- and seven trans-sQTL (Bonferroni corrected p < 0.05).” Information regarding these cis- sQTLs can be found here:https://api.gdc.cancer.gov/data/3920a044-6874-4049-8010-55f0922243b7. However, the document that provided does not indicate that these are actual splicing mutations. The document contains SNP coordinates, but not sequence changes. An assessment of known rsIDs at these coordinates only accounted for only 7 of the variants in our database. In fact, none of the substitutions (nor the rsIDs) in the article could be related back to changes in mRNA splice strength. It is not possible to comparison the results presented in this paper to those in Kahles et al. (2018). 3. The authors notice that the number of variants which activate cryptic splicing exceed the number reported in a recently published study in Cell Reports. Similarly to before: what is the extent of overlap between the two datasets? Stating that a dataset is bigger than another one is not necessarily an indication that it is better. Response: Jayasinghe et al. Cell Rep.;23(1):270-281.e3 (2018) identified 2056 variants in TCGA patients which activated cryptic splice sites, of which 1964 were confirmed by manual review of RNAseq data using the Integrated Genome Viewer (IGV). The data provided (Supplementary Table S1; “Passed Manual Review” tab) was sufficient to perform a comparison with our results. We scanned their manually reviewed variants using the Shannon Pipeline (SP). Despite reporting 1964 manually reviewed mutations, 50 mutations are shared between multiple patients (there are 1914 unique mutations total). 1510 of these mutations were found to alter at least one mRNA splice site (natural or cryptic), of which 1176 met our SP filtering criteria (either decreased natural site strength or cryptic splice sites strengthened by ≥2 bits and exceeding the strength of the nearest natural site of the same polarity). We considered the possibility that splicing mutations in Jayasinghe et al. that were not flagged by SP could have instead altered the strength of splicing regulatory factor binding sites (SRFs). The high-throughput IT-based variant analysis tools needed to address this question were not available at the time the TCGA genomic data were processed. We recently introduced a new version of SP which is capable of analyzing variants impacting both constitutive splice sites and SRFs (including SRSF1, SRSF2, SRSF5, SRSF6, hnRNPA1, ELAVL1, PTB and TIA1). Upon analysis of the variants in Jayasinghe et al. for IT-derived changes in natural, cryptic and SRF binding site strengths, 1746 of the 1914 (91.2%) were found to be significant. It is conceivable that remaining unclassified variants may affect binding by splicing factors for which we have not yet derived iPWMs (i.e. SRSF7). Of the 1176 variants meeting our filtering criteria, 824 were flagged by Veridical (70.1%). Interestingly, 27 Veridical-flagged mutations alter the same genomic coordinate in different tumours and another 64 affect the adjacent genomic coordinate within the same splice site in other individuals. We further investigated the remaining mutations (those evaluated, but not flagged) to better understand the discrepancy. In approximately 10% of cases, Veridical did not find any alternative validating cryptic splicing events in the region that contrasted with the read distribution in the set of control transcriptomes. For example, chr9:35389842G>C in TCGA-CN-5370 was expected to abolish the natural acceptor site of UNC13B exon 24, however the read counts for intron inclusion in the RNAseq data were too sparse to be deemed significant (p=0.33). Jayasinghe et al. also found mutations in TCGA patients that were not evaluated in our study. This could occur for either because the RNASeq BAM file for a particular TCGA patient failed to download, or the “key” file that associates the BAM file to its TCGA name was incomplete in some tumors (i.e. TCGA-CG-4436; TCGA-STAD), due to both the BAM file name and header lacking this information. This would not impact the accuracy of the data that is present in our database or that we report, only the level of concordance between our results and those of Jayasinghe et al. When comparing this dataset with our own, we discovered an instance where discrepant RNAseq data for the same tumour in the same TCGA patient led to different IT results. Originally, Veridical did not find a significant splicing change in POLR1B exon 14 at the +1 position, a G>A mutation g.113331138G>A, present in patient TCGA-Z6-AAPN in the ESCA dataset. When we manually reviewed the RNAseq BAM file used in this analysis, alternate or exon skipped splice forms were not observed, confirming the results reported from Veridical. In fact, TCGA deposited two separate BAM files containing RNAseq data for this same patient: “TCGA-Z6-AAPN-01A-11R-A406-31_rnaseq.bam” and “UNCID_2681450.b17c4505-8a84-4cdd-8782-fbc456deb2a6.sorted_genome_alignments.bam”. The latter BAM file shows evidence of both predicted exon skipping and activation of a cryptic acceptor site 12 nt downstream of exon 14 of POLR1B. Comparison of SNPs between these BAM files in other genes did not show any evidence of sample switching or contamination. Although this issue does not appear to be widespread in the TCGA dataset, such discrepancies exceed the scope of our study, and rightfully should be addressed by TCGA. Nevertheless, this discovery did prompt us to re-analyze the TCGA-ESCA variants through Veridical using the second set of BAM files, and we have now included those results to ValidSpliceMut. In this new set, the previously mentioned POLR1B mutation is deemed significant due to increased exon skipping (86 reads showing exon skipping; p=0.0000). 4. The authors validate ~27% of predicted splicing variants using the mRNA data (351k validated of the 1.2M predicted). This is a surprisingly low fraction. Later in the manuscript, the authors briefly discuss about the possible reasons of such a discrepancy. One of them is the possible occurrence of nonsense mediate decay which will not confirm the mutations because no or very few reads will be detected. However, as the authors acknowledge, the absence of supporting reads only in mutated individuals as compared to the presence of reads in WT sample would be a strong indication of the effective role of these mutations on splicing. This can be quantified from the same RNAseq data and in my opinion should be done. In the revision of this paper, the fraction of validated predicted splicing variants in the ValidSpliceMut database increased from 27% to 31%, as a result of a series of improvements in the software used for processing. SP was significantly upgraded over the course of this project,. The previous iteration would incorrectly report information changes at pre-existing splice sites adjacent to certain mutations. These sites were characterized by genomic coordinates that were altered by insertions or deletions (indel), regardless of whether the site overlapped or included the sequence change. In such instances, some altered natural splice sites could be designated as cryptic sites. SP also reported changes at cryptic acceptor and donor sites in the first and last exons of a gene, respectively, which were not likely to have a meaningful impact on splicing. Therefore, all datasets processed prior to this upgrade were reanalyzed for indel variants. The TCGA ESCA dataset was reprocessed with a second set of RNAseq BAM files (see above response to point 3), which increased the fraction of flagged mutations for that tumor type. Finally, we processed an additional 7 tumor datasets from ICGC and included the validated mutations in our primary beacon database. The statistics of mutations, their distributions and support have been updated in the present version of the manuscript. From our perspective, the proportion of variants validated is not “a surprisingly low fraction\". The results reported here are consistent with our previous published studies (Dorman et al., Sci. Rep. 4: 7063, 2014). Aside from NMD, as demonstrated below, some mutations that significantly alter splice site strength may not have been flagged by Veridical as a consequence of low levels of expression of the gene in the tumor (or controls) itself. Furthermore, Veridical cannot make an accurate assessment of the region of interest in control samples if these lack sufficient read abundance levels to determine the probability (p-value) of observing expression in the mutation-containing vs control samples. Also, our analysis did not take into account other impacts of the variant on sequences that influence exon definition, such as binding to splicing regulatory factors or mRNA secondary structure. In our response to point 3, we described a discrepancy in BAM file sources, which could also lead to a lower fraction of confirmed variants. Finally, miscalled variants (despite the stringent quality control criteria applied to select variants) could contribute to the fraction of variants not supported by Veridical analysis. Such technical artifacts have been shown to be quite common in exome sequencing in areas of the genome characterized by low mappability (from Shi et al., Cell Rep. 2018 Nov 6;25(6):1446-1457, 2018): “Examination of the genomic locations of mutations revealed that 41.1% of the artifactual somatic mutations occurred in regions of low mappability compared with only 6.4% for the validated somatic heterogeneous mutations.” As indicated in our response to point 3, the previous version of this paper only evaluated variants for their impact on constitutional mRNA splice sites and cryptic sites, and excluded impacts of mutations at splicing regulatory factor binding sites (SRFs). The scope and time required to assess SRFs precluded the reanalysis of all datasets for such changes. However, to address the issue raised by the reviewer, we evaluated the degree to which ignoring SRFs would affect the overall discovery of splicing-related variants. The updated version of the Shannon Pipeline (SP) with this capability was used to examine 1050 mRNA splicing variants that have been demonstrated to affect exon recognition (Cheung et al., Mol Cell. 2019 Jan 3;73(1):183-194.e8). These splicing variants were experimentally validated using a high throughput, multiplexed splicing minigene reporter assay in that study. SP reported a change in splicing and/or SRF binding strength for 1017 of these 1050 mutations (96.9%; where change in SRF strength was ≥ 3 bits). After accounting for SRF location (e.g. exonic TIA1 sites were eliminated, since these have not been proven to have splicing effects; see Table 1 of Caminsky et al., F1000Res.;3:282, 2014, for a full description of each SRF), the number of flagged variants was reduced to 940 (89.5%). Based on changes in constitutive splice site strength alone, 447 variants were flagged (435 weaken natural sites, 14 strengthen cryptic sites to a level exceeding the nearest natural site). Therefore, 46.3% of the constitutive mutations at natural or cryptic splice sites were also flagged by SP. This suggests that the lower predictive accuracy of SP in our original submission was, in part, due to the limitations in its ability to detect pathogenic mutations in SRF binding sites. We addressed the reviewer’s suggestion to compare expression of the same gene in tumours with Veridical-validated mutations with other tumors with SP-predicted mutations in the same gene that were not experimentally validated. To perform the analysis, we obtained pre-processed mRNA expression data from the same RNAseq sources of TCGA patients from cBioPortal (www.cbioportal.com; provisional datasets were used, which contained largest number of patients for each tumor type). We extracted these gene expression values with a software program we wrote that determined transcripts per million (TPM) for each gene containing a SP-flagged variant. Expression values for gene present multiple times (due to the presence of multiple splice isoforms) were averaged for the particular tissue from which they were derived. We separated mRNA expression values for each gene in TCGA patients into the Veridical-flagged vs. non-validated SP-predicted mutation categories, and performed a Student’s t-test on the two groups. The expression values of 58.2% of genes were statistically distinct with 90% confidence; >2 patients per category per gene). With at least 10 patients per category, the number of statistically different genes increased to 69.3%. Among these genes, patients with Veridical-flagged variants had higher overall gene expression in 99.7% of cases. These inherent differences in expression suggest that the failure to validate predicted mutations may be related to little or no expression of these genes in tumour and/or control samples, rather than to accuracy of IT prediction methods. Non-sense mediated decay could be responsible for the decreased expression of these mutated genes in the tumour genomes that carry them, or the failure to validate could be related to low levels of expression of these genes in the particular tissues from which the tumours were derived. This analysis is now described in the revised manuscript (‘Dataset validation and discussion,’ para. 8). Note: The gene expression data from cBioPortal had some limitations: 4196 genes containing variants flagged by SP are not present in the mRNA expression datasets, though the vast majority occurred in non-coding RNAs (i.e. 145 microRNAs, 194 LINC RNAs) or other uncharacterized RNAs (e.g. 324 ‘LOC’ RNAs). Furthermore, certain TCGA patients that we analyzed were not available in cBioPortal among the available expression datasets (2 TCGA-BRCA, 18 TCGA-COAD, 19 TCGA-GBM, 1 TCGA-HNSC, 1 TCGA-LUAD, 119 TCGA-OV, 4 TCGA-READ, 4 TCGA-STAD, 4 TCGA-THCA, and 7 TCGA-UCEC patients). Nevertheless, sufficient data were obtained for the analysis that we carried out. Furthermore, this analysis has other caveats, especially regarding the accuracy of RNAseq quantification between samples, even upon normalization of the data. The vast majority of TCGA tumor samples do not have a matched normal counterpart, and many TCGA tumor datasets (e.g. TCGA-LAML) do not provide any normal control RNAseq data at all. While we could perform an analysis in which the tumor is compared to a set of normals for the same tissue, variation in expression between different individuals would obfuscate evidence of any apparent NMD caused by the mutation. A 50% reduction of expression (or less) may not be observable under these conditions. 5. In general, the authors seem to exclude that their prediction method could lead to false positives. Rather they justify the poor overlap with limitations of mRNA detection. If this is the case, this should be quantified and probably a comparison with other prediction methods could help. Response: We have quantified RNA expression in tumors for mutations that were validated by Veridical vs those which were not (see response to point 4). Regarding false positives: An extensive comparison of Information Theory to other bioinformatic programs which evaluate variants for splicing impact (MaxEntScan and Human Splice Finder) has been performed (Mucaki et al. 2013; Caminsky et al. 2014; see response to point 1). False positives are extremely rare because strength of binding sites (in bits) is directly related to their binding affinities; we have demonstrated that unused cryptic splice sites in the vicinity of natural splice sites are significantly weaker. Based on our experience and published analyses of a genome-wide site of binding sites (Rogan et al. 2003), such decoy splice sites are nearly always at least 4 bits (2^4 or 16 fold) weaker than sites that are actually recognized by spliceosomes. The predictive accuracy of the IT methodology for detecting expression-validated mutations was determined to be 87.9% (762 of 867 variants from 122 different publications; changes to SRFs were included in this variant dataset; Caminsky et al. 2014). This value is similar to the predictions made to those of Cheung et al. (2019), where we predicted splice site and/or SRFs changes to 89.5% mutations experimentally validated to cause exon definition events. The performance of IT-based methods for predicting splicing mutations has been well established over the past two and a half decades. Re-evaluation of its accuracy is not necessary, and this issue is, at best, only tangentially relevant, to the purpose of presenting the resource described in a Data Note article. 6. Of the >351k mutations with an effect on splicing supported by RNA data, only 35 affect CGC genes. Is this only a subset of mutations affecting driver genes or is it the complete list? In the former case, I would suggest that the authors provide the full list as supplementary data. In the latter case, the authors should discuss the implication of such a low number. Considering that there are >700 CGC genes, does it mean that aberrant splicing is very rarely a driver event? Is the overwhelming majority of splicing variants passenger? Response: There are 25 CGC genes indicated in Table 2, however these were never intended to be interpreted to be a complete list of CGC genes with Veridical-flagged mutations. The table has been renamed to indicate that these are a set of representative mutations. In the previous version of this paper, the number of variants (“n=25”) did not indicate the total number of CGC genes. This has been removed and replaced with the actual number of CGC splicing mutations that were validated: “In Table 2, we highlight a subset of validated splicing mutations which were identified in known driver genes implicated in the COSMIC (Catalogue Of Somatic Mutations In Cancer) Cancer Gene Census catalog (CGC) 27. In total, 543 “Tier 1” CGC genes have at least one Veridical-flagged variant present in the ValidSpliceMut database.”" } ] } ]	1	https://f1000research.com/articles/7-1908
https://f1000research.com/articles/8-1588/v1	04 Sep 19	{ "type": "Research Article", "title": "Management and cost of snakebite injuries at a teaching and referral hospital in Western Kenya", "authors": [ "Mitchel Otieno Okumu", "Minal Naran Patel", "Foram Rajnkant Bhogayata", "Francis Okumu Ochola", "Irene Awuor Olweny", "Joshua Orungo Onono", "Joseph Kangangi Gikunju", "Minal Naran Patel", "Foram Rajnkant Bhogayata", "Francis Okumu Ochola", "Irene Awuor Olweny", "Joshua Orungo Onono", "Joseph Kangangi Gikunju" ], "abstract": "Background: Data on the cost of snakebite injuries may inform key pillars of universal health coverage including proper planning, allocation, and utility of resources. This study evaluated the injuries, management, and costs resulting from snakebites at Jaramogi Oginga Odinga Teaching and Referral Hospital (JOOTRH) in Kenya. Methods: In total, medical records of 127 snakebite victims attending JOOTRH between January 2011 and December 2016 were purposely selected and data on the age, gender, type of residence (urban or rural), part of the body bitten, time of bite, injuries, pre-hospital first aid, time to hospital, length of stay, treatment, and costs were collected. Regression analysis was used to predict the total indirect cost of snakebite injuries and p≤ 0.05 was considered significant. Mortality and loss of income of hospitalized victims were considered as direct costs. Results: It was found that 43 victims were 13-24 years of age, 64 were female, 94 were from rural areas, 92 were bitten on the lower limbs, 49 were bitten between 6.00 pm and midnight, 43 attempted pre-hospital first aid, and the median time to hospital was 4.5 hours. Antivenom, supportive therapy, antibiotics, antihistamines, corticosteroids, analgesics, and non-steroidal anti-inflammatory drugs were used. Cellulitis, compartment syndrome, gangrenous foot, psychiatric disorder, and death were the main complications. Most victims spent 1-5 days in hospital and the median cost of treating a snakebite was 2652 KES (~$26). Drugs, ward charges, and nursing procedures were the highest contributors to the total indirect cost. Victims hospitalized for 6-10 days and >10 days incurred 32% and 62% more costs, respectively, compared to those hospitalized for 1-5 days. Conclusions: The longer snakebite victims are hospitalized, the higher the cost incurred. Continuous medical education on the correct management of snakebites should be encouraged to minimize complications that may increase hospital stays and costs incurred.", "keywords": [ "snakebite", "snakebite envenoming", "costs of snakebite", "universal health coverage", "neglected tropical disease", "snakebite burden", "sub-Saharan Africa", "management of snakebite" ], "content": "Introduction\n\nSnakebite envenoming is a topic perhaps too little discussed on the global stage. The disease has only recently been formally re-instated to the WHO list of neglected tropical diseases after a four-year hiatus1. The additional focus is welcome and has been widely anticipated by many people that deal with and understand the devastating impact of snakebite envenoming on individuals, their families, and communities in general. It is widely hoped that the reinstatement will significantly boost efforts to reduce the burden of snakebite in Asia, Africa, and Latin America where the burden is highest1,2.\n\nSpecifying the requirements for antivenom at the local level, educating at-risk groups, improving accessibility to antivenoms, and training of healthcare personnel are the major challenges in addressing snakebites1. Most of the mitigative strategies focus on ways and means of addressing these challenges1. However, evaluating the cost associated with managing snakebites at the local hospital level has often been overlooked. To date, there are no studies that provide estimates for how much it may cost to treat a victim of a snakebite at the local hospital setting.\n\nKenya is a country with a population of about 40 million people3. Access to affordable healthcare in the country has often been fraught with controversy4. In a bid to provide quality healthcare to its citizens, the government of Kenya has recently adopted the big four agenda5, which seeks to deliver healthcare, food security, manufacturing growth, and affordable housing5. To deliver on healthcare, the government has embraced the Kenyan Health Policy, which borrows heavily from the WHO model on Universal Healthcare Coverage (UHC)6. The aim is to ensure that every Kenyan receives quality healthcare (promotive, preventive, curative and rehabilitative) without suffering financial strain7.\n\nThere has been a huge clamor for UHC in Kenya. However, only four counties/administrative units, namely Kisumu, Isiolo, Nyeri, and Machakos, have been selected to pilot the rollout of the program7. Data on the cost of treating various diseases may inform key pillars of UHC, including proper health planning, resource allocation and utility of resources. Snakebite injuries are probably one of the most hidden public health crises globally. However, in a country that is grappling with HIV/AIDS, malaria, tuberculosis and other non-communicable diseases8, there is a real risk that snakebites may be further neglected in terms of policy and resource allocation. Estimating the cost of managing snakebite injuries at the local hospital level may go a long way in highlighting the gaps that require the attention of the government and other relevant stakeholders.\n\nIn a previous study on acute poisoning at Jaramogi Oginga Odinga Teaching and Referral Hospital (JOOTRH), we established that snakebites were the leading cause of poisoning between the years 2011 and 20169. Therefore, the objective of the present study was to evaluate the injuries, management, and costs associated with snakebites at the hospital between the years 2011 and 2016. The information may be important for guiding policy on proper health planning, resource allocation and utility of the limited resources available to the counties.\n\n\nMethods\n\nPermission to conduct the study was obtained from the hospital administration; ethical clearance was obtained from the Ethical Review Committee (ERC) of JOOTRH (Ref no: ERC.IB/VOL1/412). The ethical approval document is provided as Extended data (Figure S1)10.\n\nThis was a descriptive cross-sectional study that was carried out between July and November 2017 using data from a retrospective audit on records of acute poisoning at JOOTRH9. All patients with acute poisoning due to snakebite envenoming presenting to and managed in the emergency department of JOOTRH between January 2011 and December 2016 were reviewed for inclusion. Victims transferred from health centers, dispensaries, and sub-county hospitals to JOOTRH were counted once after thorough scrutiny of the medical records to ensure there was no double reporting. The medical records had to have all the study variables of interest; age, gender, type of residence (urban or rural), part of the body bitten, time of bite, injuries, pre-hospital first aid, time to hospital, length of stay, treatment, and costs. Any medical records that had incomplete information or did not have any of these parameters of interest were excluded from the analysis. In total, 127 medical records of snakebite victims attending the hospital between January 2011 and December 2016 were purposely selected for the study. The digital archiving of medical records at JOOTRH based on the international classification of diseases begun in January 2011 and this made it easier for us to access/trace the physical medical records, as opposed to periods before the digital archiving was in place. We therefore selected a five-year time frame from the period of inception of the digital archiving as our study period. Data on the age, gender, type of residence (urban or rural) of the victim, part of the body bitten, and the time of the bite was retrieved. Other data retrieved included the type of injuries, pre-hospital management measures, time taken to reach the hospital, length of stay at the hospital, type of treatment, and associated costs. Indirect costs included those related to registration, admission, file fees, laboratory, laundry, physiotherapy, nursing procedures, antivenom, other drugs, daily ward charges, caretaker costs, cost of non-pharmaceuticals, surgical operations on victims, and miscellaneous. The direct costs considered in this analysis were those incurred due to mortalities and loss of income by patients who were hospitalized.\n\nThe study was carried out at JOOTRH, a referral hospital in Kisumu County. In December of 2018, the county was selected as the first of four counties to pilot UHC in Kenya7. The hospital has a wide catchment area encompassing up to 10 counties (Kisumu, Siaya, Homa Bay, Migori, Kisii, Kakamega, Vihiga, Bungoma, Busia, and Nandi) within the Western Kenya region. Based on estimates from the most recent census (2009), these counties are estimated to have a population of around 970, 000, 840,000, 750,000, 260,000, 1,150,000, 1,660,000, 550, 000, 1,600,000, 490, 000, and 750,000 people, respectively9. Curative, preventive, promotive and rehabilitative health services are provided at JOOTRH9.\n\nSociodemographic data from the pre-structured proforma were collated in MS Excel spreadsheets and analyzed using descriptive statistics. Categorical variables were presented in frequencies and percentages and qualitative variables were handled through thematic analysis. Quantitative data including additional costs of treatment were analyzed using descriptive statistical measures including measures of central tendency; mean, median and measures of dispersion: minimum, and maximum. Multiple linear regression was used to determine the predictors of the total indirect cost of treating snakebite (SPSS version 20.0). The total indirect cost (dependent variable) was transformed to log10 and regressed against independent variables including the age, gender, and type of residence (urban or rural) of victims, location of the victims at the time of the bite, season and time of the bite, and the time taken to get to the hospital. Other independent variables were the length of hospital stay, first aid measures initiated, and part of the body bitten. p<0.05 was considered significant. The direct costs considered in this analysis were those incurred due to mortalities and loss of income by patients who were hospitalized.\n\n\nResults\n\nThe year of the bite, the type of residence (urban or rural) of victims and the seasonal distribution of snakebite in the catchment area of the hospital is summarized in Table 111. Most bites occurred in the year 2012, with Kisumu County recording the highest number of bites. Moreover, most bites took place during the rainy season.\n\nThe majority of the victims (110/127, 86.6%) presented to hospitals within Kisumu County, 8/127 (6.3%) cases presented to hospitals in Siaya County, 4/127 (3.1%) cases presented to hospitals in Vihiga County, 2/127 (1.6%) cases presented to hospitals in Homabay County, one (0.8%) case presented to a hospital in Busia County and another one (0.8%) case presented to a hospital in Migori County. Of the 127 cases, 84 (66.1%) presented to JOOTRH as the first port of call. This was followed by nine cases (7.1%) who presented to Ahero Sub-County hospital, four cases (3.1%) who presented to the Kombewa Sub-County hospital, and another four cases (3.1%) who presented to the Nyakach Sub-County hospital (Table S1, Extended data)10. The gender, type of residence, marital status, and age of the victims who presented to JOOTRH during the study period are summarized in Table 2. About 94/127 (74.0%) of all snakebite cases were from the rural areas and most victims were between 13 and 24 years of age.\n\nThe occupation of snakebite victims is summarized in Table 3.\n\nThe site of the bite, location of the victim at the time of the bite, and the circumstance/activity during which the snakebite occurred are shown in Table 4. Most bites were on the lower limbs, occurred outdoors, and took place while the victims were walking.\n\nMost bites occurred between 1800 and 2359 and a majority of the victims did not attempt any form of pre-hospital measure after being bitten (Table 5).\n\nMost of the victims took less than six hours to present to the hospital (median time was 4.5 hours). A majority reported having been bitten by what they described as a black snake and cellulitis was the most common complication (Table 6).\n\nThe symptoms of victims of snakebite who presented to JOOTRH during the study period are summarized in Table 7. The symptoms were local and systemic (neurological, hematological).\n\nThe treatment received by the victims of snakebite included the use of antivenom, supportive therapy, antimicrobial agents, antihistamines, corticosteroids, non-steroidal anti-inflammatory agents, opioid and non-opioid analgesics and general anesthetics (Table 8).\n\nIV: intravenous, PO: per oral, IM: intramuscular.\n\nOf the 127 victims of snakebite, 53/127 (42%) received antivenom while 74/127 (58%) did not. Of those that received antivenom, 51 survived while two died. Among the victims who received antivenom, 44 received one vial of antivenom, six received two vials of antivenom, one received three vials of antivenom, and another two received five vials of antivenom (Table S2, Extended data)10. Among those who did not receive antivenom, 72 survived and two died.\n\nOf the 52 victims of snakebite who survived after receiving antivenom, 15 had complications including cellulitis (n=7), compartment syndrome (n=4), gangrenous foot (n=2), a psychiatric episode (n=1), and soft tissue injury (n=1). None of the 74 victims who did not receive antivenom developed any complications.\n\nThe majority (107/127, 84%) of all victims of snakebite spent between one and five days in the hospital, 12 (9%) spent between six and 10 days, two (1.6%) spent between 11 and 15 days and another two (1.6%) spent between 16 and 20 days (Figure 1). There was one victim (0.8%) who spent between 21 and 25 days in the hospital, while another two (1.6%) victims spent between 26 and 30 days in the hospital (Figure 1).\n\nWhen considered in terms of the daily minimum wage in Kenya (KES 452; ~ $4 USD), victims of snakebite who spent at least five, 10 and 20 days in the hospital lost about KES 2260 (~$22), KES 4520 (~$44) and KES 9, 040 (~$88) worth of wages, respectively.\n\nThe total indirect cost of managing snakebite in the study area during the period of study was KES 568,557.72 (~$5530). Seven victims of snakebite received a waiver on the total indirect costs amounting to 91,601 KES (~$890). None of these victims had any hospital insurance cover (Table S3, Extended data)10. The highest contributors to the total indirect cost of snakebite were drugs (KES 152,964; ~$1485), ward charges (KES 142,300 KES; ~$1380), and nursing procedures (KES 81,500; ~$790) (Figure 2).\n\nThe median cost of treating snakebite was KES 2652 (~$26; range: KES 1100-41399 or ~11-400$). The predictors of the total indirect cost of treating snakebite are summarized in Table 9. Generally, victims who spent 6–10 days and >10 days incurred 32% and 62% more costs, respectively, compared to those who spent 1–5 days.\n\n\nDiscussion\n\nTo the best of our knowledge, this is the first study that has reported on the cost of snakebites in any hospital setting. Our findings suggest that drugs (excluding antivenom), ward charges, and nursing procedures are the highest contributors to the total indirect cost of managing snakebite. When antivenom is added to the valuation, the problem of managing snakebite in a local hospital setting suddenly becomes more complex. We also report that the median cost of treating snakebite was 2652 KES (~$26). According to a recently published survey, nearly half of Kenyan households earn less than 10,000 KES (~$97) per month, while 2% have no income at all12. What this means is that, in the event of a snakebite, around 50% of all Kenyan households would need to spend approximately 25% of their monthly income treating the condition, and another 2% would have no means of paying for the treatment. Moreover, the minimum wage in Kenya is currently KES 13,572 (~$134) per month, which translates to about KES 452 (~$4) per day13. Hence, for a person who misses work for 10 days due to hospitalization, the lost revenue would be about KES 4,520 (~$44). In essence, this person becomes poorer as they may not be able to buy food for their family, pay rent, or pay school fees for their children. Their productivity is therefore dented, which is bound to affect their socioeconomic standing. Moreover, the four individuals in this study who died from their injuries may also have an impact on cost, as these individuals will no longer be financially available for their families who are most likely to slide into poverty.\n\nFrom our findings, the longer a snakebite victim was in the hospital, the more likely they were to incur higher costs. Additionally, we established that spending more than 10 days in the hospital was associated with a 62% increase in the total cost, relative to spending between one and five days in hospital. A length of stay of between six and 10 days was associated with a 32% increase in the total cost of treating the snakebite at the hospital.\n\nThe first guidelines for the prevention, diagnosis, and management of snakebite envenoming in Kenya were published in April 20196. These guidelines provide details of how snakebite envenoming should be managed at the different levels of care in Kenya, including at the rural dispensary/health center, sub-county/district hospital, and at the referral hospital level6. Based on our findings, there was no evidence that the 20-minute whole blood clotting test or any urine exam, considered as standard diagnostic tests in snakebite guidelines, were done in any of these facilities. Moreover, non-steroidal anti-inflammatory drugs (NSAIDs) such as diclofenac, ketorolac, aceclofenac, and ibuprofen were routinely given to victims of snakebite, despite their use being contraindicated in managing the disease. It was also evident that some of the bites had hematological effects, including slight/mild/minimal bleeding, severe/excessive bleeding, hyper-pigmentation at the site of the bite, ecchymosis, and erythema. NSAIDs interfere with the blood coagulation cascade and may potentiate hematological disturbances in the event of a snakebite, particularly those from snakes with hemotoxic venom6. It is interesting to note that all the victims who died had received an NSAID (intramuscular diclofenac in particular) for snakebite-associated pain.\n\nAccording to the clinical notes (particularly those that had something to do with the referral chain), the unavailability of antivenom was the main reason for referring victims from the smaller facilities to either the sub-county/district hospital or to the referral hospital. At the sub-county/district hospital level, conservative management of snakebite by use of blood transfusion or fresh frozen plasma (FFP) was often not possible as the blood/FFP was seldom available at the facilities. Furthermore, sub-county/district hospitals had to refer patients to JOOTRH for an array of reasons, including the unavailability of antivenom, the lack of the pre-requisite reagents to perform clinical and laboratory examinations, and the lack of ECG/radiography services, as well as a lack of snakebite management skills.\n\nAt JOOTRH, antivenom was also not always available, and there were many cases of victims being asked to buy the antivenom from private pharmacies in the town of Kisumu. This was a challenge, especially for cases that occurred at night. Many of the private pharmacies in the area do not operate at night and most did not stock antivenom, further delaying treatment. Furthermore, based on our findings, there was no evidence that bacterial cultures were used to inform the choice of antibiotics for managing snakebite-associated infections. This observation is troubling considering the current state of antimicrobial resistance in developing countries including Kenya. The fact that very few victims were subjected to laboratory evaluation is also quite alarming. The reasons for this observation need to be elaborated.\n\nDespite all the shortcomings in managing snakebite at JOOTRH, there were a few positives. First, based on the two cases of gangrene that were resolved successfully (after fasciotomy), it seems that the surgical team at the facility is well equipped to manage venomous snakebite-induced necrosis. Second, it was good to see that rehabilitative services were offered to some victims by the department of physiotherapy and occupational therapy. Third, the availability of psychological services to snakebite victims was also commendable. This was exemplified by the case of a 16-year-old female victim of snakebite who had a psychiatric episode secondary to the bite, who was discharged three days after presenting to JOOTRH having received antivenom, supportive care (IV fluids) and psychological counseling.\n\nKisumu East sub-county contributed at least four in every ten cases of snakebite at JOOTRH during the study period. The area has several wards including Central Kolwa, East Kajulu, West Kajulu, East Kolwa, and Manyatta wards14. JOOTRH is also located in Kisumu East. We therefore posit that the numbers of patients presenting to the hospital from this region may be related to the proximity of the victims to the hospital.\n\nThere were more bites during the rainy season than the dry season, in agreement with other similar studies15,16. This observation may have something to do with the fact that snakes seek dry and safe shelter whenever heavy rains disrupt their habitats17.\n\nOur findings of more bites occurring in the rural areas, predominantly affecting the young people, mostly being inflicted on the lower limbs and occurring in the late evenings are consistent with the reports of other similar studies18–22.\n\nThe demographic that was most affected by snakebite included students and individuals partaking in outdoor activities. Moreover, most bites occurred while the victims were walking. In rural Africa, more often than not, students have to travel long distances to have access to education. We posit that these students may have been bitten in the evening hours as they walked back home from school.\n\nThe symptoms of snakebite envenoming and the description of the offending snakes are largely consistent with the type of snakes known to be in the area. These snakes are mostly of the Viperidae and Elapidae families6 and include the puff adder (Bitis arietans), gaboon viper (Bitis gabonica), rhinoceros viper (Bitis nasicornis), black mamba (Dendroaspis polylepis), Jameson’s mamba (Dendroaspis jamesoni), the eastern forest cobra (Naja subfulva), and the gold tree cobra (Pseudohaje goldi)6.\n\nMore than half of all the victims did not attempt any pre-hospital first aid measures. A similar number reported to the hospital within six hours of having been bitten. When taken collectively, this may suggest that the locals recognize that snakebite is a medical emergency requiring urgent medical intervention. On the other hand, the lack of initiative by the victims or their proxies in attempting any pre-hospital measure may suggest that the population is not knowledgeable on the appropriate steps to take in the event of snakebite. Where some form of pre-hospital measure was used, the measure adopted cannot be considered as beneficial. Some of the measures are contraindicated and have been shown to do more harm than good23. The use of tourniquets, incisions, suction, heat, ice, alcohol, and electric shock have all been reported to be counterproductive in snakebites23. The jury may still be out on the role of herbal medicine in snakebites. Those that oppose the practice argue that seeking treatment from traditional herbal medicine practitioners often delays access to proper medical intervention and may result in complications24,25. In contrast, proponents of herbal medicine argue that the purpose of herbal medicine is not to replace antivenom, but serve an adjunctive role, particularly in managing local effects of envenomation such as necrosis, as has been reported by several authors26–28. The latter seems to be buoyed by the fact that some phytochemicals isolated from medicinal plants have shown some promising in vitro and in vivo neutralization capacity against phospholipase A2 and metalloproteases, which are enzymes associated with local tissue damage26–29.\n\nIt is difficult to see how other practices such as burning matchsticks at the site of the bite, the use of a cloth impregnated with charcoal, the application of petroleum jelly at the site of the bite, the use of potassium permanganate and povidone-iodine, and pouring paraffin at the site of the bite could mitigate against snakebite envenoming. There is a need for public health awareness programs aimed at dissuading such harmful practices from being advanced in the management of snakebite among this population.\n\nThe cost of medicines and health services in Kenya are yet to be standardized and as such the indirect costs we have provided are mere estimates and may be higher or lower in other hospital settings. Moreover, this study did not capture information on the victims of snakebite who may have died on their way to the hospital or those who sought treatment outside the hospital’s catchment area. Furthermore, owing to the retrospective nature of this study, it was not possible to capture information on other costs incurred by victims, such as the costs incurred in transporting the victims to the hospital.\n\n\nConclusions\n\nSnakebite injuries contribute significantly to medical costs in the hospital setting. The longer snakebite victims stay in hospital, the higher the cost. Continuous medical education on the correct management of snakebites should be encouraged to minimize snakebite-related complications that may increase hospital stay and, consequently, the cost incurred by victims. Prospective work is needed to provide better estimates of the direct and indirect costs of snakebite injuries in the hospital setting.\n\n\nData availability\n\nFigshare: Raw data on the study titled ‘management and cost of snakebite injuries at a teaching and referral hospital in Western Kenya. https://doi.org/10.6084/m9.figshare.9642005.v211\n\nThis project contains the following underlying data:\n\n- Snake bite injuries-working data -modified.xlsx (raw demographic, medical and cost data for each participant)\n\nFigshare: Extended data on the study titled ‘management and cost of snakebite injuries at a teaching and referral hospital in Western Kenya’. https://doi.org/10.6084/m9.figshare.9204773.v510\n\nThis project contains the following extended data:\n\n- Figure S1: Ethical approval document from the ethical review committee of Jaramogi Oginga Odinga Teaching and Referral Hospital\n\n- Table S1: A summary of the hospitals that referred victims of snakebite to JOOTRH during the study period\n\n- Table S2: A summary of the utility of antivenom among victims of snakebite presenting to JOOTRH during the study period\n\n- Table S3: A summary of the waivers received by some victims of snakebite at JOOTRH during the study period\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "Acknowledgments\n\nThe authors acknowledge the assistance of officers at the medical records department (Jaramogi Oginga Odinga Teaching and Referral Hospital) notably the chairman Mr. Tom Kayago Morike, and others including Mr. Shem Kojo Odero, Mr. Michael Onyango Muhoma, and Ms. Miriam Akoth Okuta. We also wish to thank the administration of the Jaramogi Oginga Odinga Teaching and Referral Hospital for all the help accorded to us in collecting the data.\n\n\nReferences\n\nChippaux JP: Snakebite envenomation turns again into a neglected tropical disease! J Venom Anim Toxins Incl Trop Dis. 2017; 23: 38. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKasturiratne A, Wickremasinghe AR, De Silva N, et al.: The global burden of snakebite: a literature analysis and modelling based on regional estimates of envenoming and deaths. PLoS Med. 2008; 5(11): e218. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCensus 2009 Summary of Results Archives. Kenya National Bureau of Statistics. [cited 2019 Jul 22]. Reference Source\n\nDid NHIF trade in money meant to save mothers? Daily Nation. [cited 2019 Jul 22]. Reference Source\n\nBudget allocates Sh451bn for Big Four agenda. Business Daily. [cited 2019 Jul 22]. Reference Source\n\nTropical N, Program D: Guidelines for Prevention Diagnosis and Management of Snakebite Envenoming in Kenya Neglected Tropical Diseases Program Ministry of Health Government of Kenya. 2019; Reference Source\n\nLAUNCH OF THE UNIVERSAL HEALTH COVERAGE PILOT. [cited 2019 Jul 18]. Reference Source\n\nCDC Global Health - Kenya. [cited 2019 Jul 22]. Reference Source\n\nOkumu MO, Patel MN, Bhogayata FR, et al.: Acute Poisonings at a Regional Referral Hospital in Western Kenya. Trop Med Infect Dis. 2018; 3(3): pii: E96. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOkumu M, Patel M, Bhogayata F, et al.: Extended data on the study titled 'management and cost of snakebite injuries at a teaching and referral hospital in Western Kenya'. 2019. http://www.doi.org/10.6084/m9.figshare.9204773.v5\n\nOkumu M, Patel M, Bhogayata F, et al.: Raw data on the study titled ‘management and cost of snakebite injuries at a teaching and referral hospital in Western Kenya'. 2019. http://www.doi.org/10.6084/m9.figshare.9642005.v2\n\nSurvey reveals monthly income for Kenyan households. Elections 2017. [cited 2019 Jul 18]. Reference Source\n\nKenya Minimum Wages \| 2019 \| Data \| Chart \| Calendar \| Forecast \| News. [cited 2019 Jul 30]. Reference Source\n\nCommission E: Electoral Commission of Kenya. Statistics. 2007; 1–8.\n\nPhillips C, Lipman GS, Gugelmann H, et al.: Snakebites and climate change in California, 1997-2017. Clin Toxicol (Phila). 2019; 57(3): 168–74. PubMed Abstract \| Publisher Full Text\n\nRahman R, Faiz MA, Selim S, et al.: Annual incidence of snake bite in rural bangladesh. PLoS Negl Trop Dis. 2010; 4(10): e860. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWhy Snakes Come To Your Home During Storms. Critter Ridder Texas. [cited 2019 Jul 18]. Reference Source\n\nChafiq F, El Hattimy F, Rhalem N, et al.: Snakebites notified to the poison control center of Morocco between 2009 and 2013. J Venom Anim Toxins Incl Trop Dis. 2016; 22: 8. PubMed Abstract \| Free Full Text\n\nOliveira Nda R, Sousa AC, Belmino JF, et al.: The epidemiology of envenomation via snakebite in the State of Piauí, Northeastern Brazil. Rev Soc Bras Med Trop. 2015; 48(1): 99–104. PubMed Abstract \| Publisher Full Text\n\nFeitosa EL, Sampaio VS, Salinas JL, et al.: Older age and time to medical assistance are associated with severity and mortality of snakebites in the Brazilian Amazon: A case-control study. PLoS One. 2015; 10(7): e0132237. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWHO: Prevalence of snakebite envenoming. WHO. [cited 2019 Jul 10]; 2017. Available from: Reference Source\n\nOchola FO, Okumu MO, Muchemi GM, et al.: Epidemiology of snake bites in selected areas of Kenya. Pan Afr Med J. 2018; 29: 217. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBoyd JJ, Agazzi G, Svajda D, et al.: Venomous snakebite in mountainous terrain: prevention and management. Wilderness Environ Med. 2007; 18(3): 190–202. PubMed Abstract \| Publisher Full Text\n\nKularatne AM, Silva A, Maduwage K, et al.: Victims’ response to snakebite and socio-epidemiological factors of 1018 snakebites in a tertiary care hospital in Sri Lanka. Wilderness Environ Med. 2014; 25(1): 35–40. PubMed Abstract \| Publisher Full Text\n\nMichael GC, Thacher TD, Shehu MI: The effect of pre-hospital care for venomous snake bite on outcome in Nigeria. Trans R Soc Trop Med Hyg. 2011; 105(2): 95–101. PubMed Abstract \| Publisher Full Text\n\nMolander M, Staerk D, Mørck H, et al.: Investigation of skin permeation, ex vivo inhibition of venom-induced tissue destruction, and wound healing of African plants used against snakebites. J Ethnopharmacol. 2015; 165: 1–8. PubMed Abstract \| Publisher Full Text\n\nMolander M, Nielsen L, Søgaard S, et al.: Hyaluronidase, phospholipase A2 and protease inhibitory activity of plants used in traditional treatment of snakebite-induced tissue necrosis in Mali, DR Congo and South Africa. J Ethnopharmacol. 2014; 157: 171–80. PubMed Abstract \| Publisher Full Text\n\nLiu Y, Staerk D, Nielsen MN, et al.: High-resolution hyaluronidase inhibition profiling combined with HPLC-HRMS-SPE-NMR for identification of anti-necrosis constituents in Chinese plants used to treat snakebite. Phytochemistry. 2015; 119: 62–9. PubMed Abstract \| Publisher Full Text\n\nKrishnan SA, Dileepkumar R, Nair AS, et al.: Studies on neutralizing effect of Ophiorrhiza mungos root extract against Daboia russelii venom. J Ethnopharmacol. 2014; 151(1): 543–7. PubMed Abstract \| Publisher Full Text" }	[ { "id": "55305", "date": "12 Nov 2019", "name": "Benjamin Wachira", "expertise": [ "Reviewer Expertise Emergency Care in Kenya" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article is well written and easy to follow.\nThere is a need to classify all the 127 bite victims into the 3 categories Cytotoxic, Haematotoxic and Neurotoxic and any combinations of the same seen so it's clear how many patients had what symptoms.\nSubsequently, the patients should be followed down along these pathways so it's clear for each category. It's hard to tell if the 74 who did not get antivenom, did they have any symptoms? They may not have had toxic snake bites or were these excluded from the study...not clear.\nAs all the data is available, we should be able to track the different patient categories through their hospital stays and also determine the outcomes for each of the categories.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "55940", "date": "21 Nov 2019", "name": "Charles F.B. Nhachi", "expertise": [ "Reviewer Expertise Toxicology( including snake bite toxiclogy", "Pharmacogenetics" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a very interesting and important subject to interrogate especially in the Sub Saharan region where snake bite is a significant public health concern even though it is a neglected one. The financial burden of snake bite injuries on the health sector is even more neglected despite the fact that it is significant. So this study is indeed a welcoming breath of fresh air so to speak.\n\nThis work is acceptable for indexing with some minor corrections/revisions. While most of the current citations were referred to the paper can be enriched by citing work which has been carried out in the region, e.g. South Africa and Zimbabwe, to emphasise the commonality of the problem in the region.\n\nThe authors could also add references from the WHO “Guidelines for the Prevention and Clinical Management of Snakebites in Africa”. Although the guidelines are not very recent, they are still the prevailing ones.\n\nWhat is the age group of “child” according to the authors?\nWhat is the relevance of marital status to snakebite injuries?\n\nCircumstances of injury, “walking”, can this be explained further…walking where?\nWhat is “homemaking”?\n\nPerhaps the length of hospitalisation, the site of injury and the kind of snake can be associated to give a more vivid picture. Also the association of cause of death, snake type and injury can be elaborated.\n\nSnake identification can be confusing and difficult. A black snake is not necessarily a black mamba. A black mamba is in fact not black! It's black inside the mouth.\n\nWhat was the cause of death of the two who died after receiving antivenom? Was it due to reaction to the antivenom or late medical attention and what was the kind of injury?\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "56985", "date": "03 Dec 2019", "name": "Godpower Chinedu Michael", "expertise": [ "Reviewer Expertise snakebite", "health economics", "primary care", "malaria", "family medicine training" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nComments\nGeneral comment:\nThank you inviting me to review this manuscript. The title of this manuscript is apt and is of relevance in many resource-limited settings.\nSpecific comments:\nIntroduction: this beautifully written.\n\nMethods: Here, I appreciate the description of the study setting and reason for the choice of the period to which medical charts were reviewed (in this case, 5 years). However, I will suggest a review of your definition of costs in the economic evaluation of snakebite injuries. Generally and broadly there are two types of costs:\n\nDirect cost: Refers to the resource consumed in providing health care interventions. It consist of:\nDirect medical cost: cost of hospital stay, outpatient visits, card/consultation fees, drugs, labs, surgery, physiotherapy, etc\nDirect non-medical cost: these resources supporting medical services such as transportation cost, caregiver costs, etc.\n\nIndirect cost: this refers to cost from production losses. These include cost due to incapacity to work, disability or premature death.\n\nIf “complications including death” from snakebite were your outcome measures, then the nomenclature should be reworded to reflect these, instead of “direct cost” as it is presently.\n\nData handling and statistical analysis: I struggled to understand what qualitative variables were analyzed through thematic analysis. Your study was of quantitative design; for instance, no focused group discussions were held. This area needs clarification or else the statement I am referring to should be deleted.\n\nThe results may need to reflect the queries above.\n\nConclusion: reasonably written.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "55941", "date": "05 Dec 2019", "name": "Oommen Oommen", "expertise": [ "Reviewer Expertise Working on herbal products for snake bite treatment and to improve the antivenom quality." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI have read the above manuscript with interest and found it to be a good survey followed by their conclusions. The manuscript is well written with the data available and help understand the lacunae in snake bite management in Kenya. The study included authentic data collected for 5 years from JOOTRH between 2011-2016. They have screened 127 medical records having full desired data and made the analysis.They followed international classification from medical archives to arrive at their conclusions. The snake bite still remains a neglected tropical disease in Kenya as is similar in other tropical countries. The fact that snake bite victims have to travel a long distance to avail antivenom is certainly a big draw back. Even in some cases, the antivenom is still not available in place like JOOTRH. The results of the present study should be made available to Health officials to take adequate measures for future snake bit management. Ideally local county hospitals should be empowered to reduce travel time and cost of treatment. Antivenom may be made available in referral hospitals to reduce, mortality and cost of treatment especially when the average monthly income of citizen is $44. The quality of antivenom also to be improved, probably with support from UN. It is welcome to take initial traditional and dependable herbal medicines and avail anitvenom from nearest county hospitals.The authors deserve credit for their commitment and concern for snake bite cases in Kenya. The manuscript is well written and the language is good. I recommend the manuscript for indexing in your journal and also with a request to implement the important findings of this study to reduce the snake bite deaths and sufferings of the survivors.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? No source data required\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1588
https://f1000research.com/articles/7-1603/v1	04 Oct 18	{ "type": "Research Article", "title": "Predicting transcription factor binding using ensemble random forest models", "authors": [ "Fatemeh Behjati Ardakani", "Florian Schmidt", "Marcel H. Schulz" ], "abstract": "Background: Understanding the location and cell-type specific binding of Transcription Factors (TFs) is important in the study of gene regulation. Computational prediction of TF binding sites is challenging, because TFs often bind only to short DNA motifs and cell-type specific co-factors may work together with the same TF to determine binding. Here, we consider the problem of learning a general model for the prediction of TF binding using DNase1-seq data and TF motif description in form of position specific energy matrices (PSEMs). Methods: We use TF ChIP-seq data as a gold-standard for model training and evaluation. Our contribution is a novel ensemble learning approach using random forest classifiers. In the context of the ENCODE-DREAM in vivo TF binding site prediction challenge we consider different learning setups. Results: Our results indicate that the ensemble learning approach is able to better generalize across tissues and cell-types compared to individual tissue-specific classifiers or a classifier applied to the data aggregated across tissues. Furthermore, we show that incorporating DNase1-seq peaks is essential to reduce the false positive rate of TF binding predictions compared to considering the raw DNase1 signal. Conclusions: Analysis of important features reveals that the models preferentially select motifs of other TFs that are close interaction partners in existing protein protein-interaction networks. Code generated in the scope of this project is available on GitHub: https://github.com/SchulzLab/TFAnalysis (DOI: 10.5281/zenodo.1409697).", "keywords": [ "ENCODE-DREAM in vivo Transcription Factor binding site prediction challenge", "Transcription Factors", "Chromatin accessibility", "Ensemble learning", "Indirect-binding", "TF-complexes", "DNase1-seq" ], "content": "Introduction\n\nTranscription Factors (TFs) are key players of transcriptional regulation. They are inadmissible to maintain and establish cellular identity and are involved in several diseases1. TFs bind to the DNA at distinct positions, mostly in accessible chromatin regions2, and regulate transcription by recruiting additional proteins. The TFs can alter chromatin organization or, for example, recruit an RNA polymerase to initiate transcription1. Hence, to understand the function of TFs it is vital to identify the genomic location of TF binding sites (TFBS). As TFs regulate distinct genes in distinct tissues, these binding sites are tissue-specific2.\n\nNowadays, the most prevalent and widely used method to experimentally determine TFBS is through ChIPseq experiments, which can be used to generate genome-wide, tissue-specific maps of in-vivo TF binding. However, ChIP-seq experiments are expensive, experimentally challenging, and require an antibody for the target TF. To overcome these limitations, a number of computational methods have been developed to pinpoint TFBS. Most of these methods are based on position weight matrices (PWMs) describing the sequence preference of TFs3. PWMs indicate, for each position of a TF binding motif independently, which nucleotide is most likely to occur. Unfortunately, screening the entire genome using a PWM results in too many false positive predictions. Therefore, numerous methods have been proposed to reduce the prediction error by combining PWMs with epigenetics data, such as DNase1-seq, ATAC-seq, or Histone Modifications, reflecting chromatin accessibility. Also, additional features such as nucleotide composition, DNA shape, or sequence conservation can be incorporated into the predictions. Including these additional data sets and information improved the TF binding predictions considerably4–11. A non-exhaustive overview is provided in 12. While PWM based models are still the most common means to assess the likelihood of a TF binding to genomic sequences, more elaborate approaches such as SLIM-models, which capture nucleotide dependencies, have been successfully used as well13. Recently, deep learning methods have been used to learn TF binding specificities de novo from large scale data sets comprising not only ChIP-seq but also Selex and protein binding microarray (PBM) data14.\n\nThe ENCODE-DREAM in vivo Transcription Factor binding site prediction challenge15 aims to systematically compare various approaches on TFBS prediction in a controlled setup, with the additional complexity of applying the classifiers on the tissues/cell types that were not used for model training. The challenge organizers provide TF-ChIP seq data for 31 TFs, accompanied with RNA-seq and DNase1-seq data in 12 different tissues. Using labels deduced from the TF-ChIP-seq data, predictive models for TF binding should be learned and then applied to a set of hold-out chromosomes on an unseen tissue. Predictions are computed in bins, covering the entire target chromosomes. The main challenge paper will provide a detailed explanation of the challenge setup and a comparison across all competing methods. This article is a companion paper to the main ENCODE-DREAM Challenge paper, in which we describe our contribution to the challenge, delineate the motivation for our work and provide an independent evaluation of our ideas to achieve generalizability across tissues.\n\nWe developed an ensemble learning approach using random forest (RF) classifiers, extending the work of Liu et al.11. Tissue-specific cofactor information was shown to be relevant to accurately model TF binding11,16. Thus, we designed our approach to aggregate tissue-specific cofactor data, via an ensemble step, into a generalizable model. Briefly, we compute TF affinities with TRAP17 for 557 PWMs in DNase-hypersensitive sites (DHSs) identified with JAMM18. TF affinities computed by TRAP are inferred from a biophysical model. In contrast to a simple binary classification, e.g. FIMO19, these scores can capture low affinity binding sites, which were shown to be biologically relevant20,21. Here, we show that our ensemble models generalize well between tissues and that they exhibit better classification performance than tissue-specific RF classifiers. Furthermore, we illustrate that only a small subset of TF features is sufficient to predict tissue-specific TFBSs and also show that these TFs are often known co-factors/interaction partners of the target TF.\n\n\nMethods\n\nWithin the scope of the challenge participants were provided with ChIP-seq data for 31 TFs, as well as DNase1-seq and gene expression obtained from RNA-seq data for 13 tissues. From the available 31 TFs, 12 were used to assess the model performance in the final round of the challenge. Hence, we also focus on these 12 TFs in the scope of this article: CTCF, E2F1, EGR1, FOXA1, FOXA2, GABPA, HNF4A, JUND, MAX, NANOG, REST, and TAF1. The number of binding sites per TF and tissue is shown in Table 1. Note that we exclude ambiguous sites from consideration in this study. We refer to the challenge website for a detailed overview on the provided data15. The challenge required that the predictions are made in bins of size 200bp, shifted by 50bp each, spanning the whole genome.\n\nIn order to obtain datasets per tissue and per TF that could be handled in terms of memory consumption and processing time, and also to cope with the large imbalance number of bound and unbound sites, we randomly sampled as many negative sites from the provided ChIP-seq tsv files as there were true binding sites per TF. The ChIP-seq labels contained in the balanced and down sampled tsv files are used as the response for training RF models.\n\nThroughout the course of challenge, we have used two distinct ways to generate features for the RF classifiers: (1) with and (2) without considering DHSs. In none of the approaches have we used the provided RNA-seq data nor did we compute DNA shape features. Generally, we computed TF binding affinities with TRAP17 for 557 distinct TFs using the default parameter settings. The position specific energy matrices (PSEMs) used in our computation are converted from position weight matrices (PWMs) obtained from JASPAR3, UniPROBE22, and Hocomoco23. The code to perform the conversion and to run TRAP is available on GitHub.\n\nWe compared two approaches to generate features for the classifier from DNase1-seq data. In the first approach, shown in Figure 1a, we compute tissue-specific DHSs using the peak caller JAMM18 (version 1.0.7.2) and merge the peak calls using the bedtools merge24 command (bedtools version 2.25.0) . Next, TF affinities are calculated in the identified DHS sites using TRAP, and the median DHS signal per peak is computed from the provided bigwig files. The computed data is intersected, using a left outer join with bedtools, with the binned genome structure required for training (using the bins contained in the tsv files mentioned above) and testing (using the provided bed-file containing all test regions).\n\n(a) Data preprocessing workflow using DNase1 Hypersensitive Sites (DHSs). Using JAMM, DHSs are called considering all available replicates for a distinct tissue. Transcription factor (TF) affinities in the identified DHSs are computed using TRAP for 557 TFs, the median signal of DHSs is assessed using bedtools. (b) An alternative data preprocessing workflow without DHSs: TF affinities and median DNase1-seq signal are computed per bin.\n\nThe second approach for computing the features is depicted in Figure 1b. Here, we do not use the information on DHS sites, instead we compute TF binding affinities and the DNase1-seq signal per bin. To account for variability between both biological and technical replicates, we calculate the median DNase1 coverage across the replicates using the bedtools coverage command. Overall, the features for a single bin are composed of the TF affinities in that bin, the DNase1 signal in the bin itself together with its left and right neighboring bins.\n\nThe Random Forest models, implemented using the randomForest R-package25 (version 4.6-12), are trained on either of the feature setups explained in the previous section. Training the RF models can be seen as a two step approach that is independent from the feature setup. Throughout model training, the balance between the bound and unbound classes is maintained to avoid over-fitting of the RF classifiers and also to ensure an unbiased evaluation of model performance. For fitting the RF classifiers we used 4,500 trees, and at most 30,000 positive and negative, i.e. bound and unbound, samples. This restriction is enforced by the limitations of the randomForest R-package. As illustrated in Figure 2a, for a given target TF, we first learn tissue and TF specific RF classifiers using all available features from the input matrix, Ti ∈ Rn×557 ; i ∈ {1, ... ,m}, where n is the number of bins forming the training set, and m denotes the number of training tissues for the target TF:\n\n\n\nhere Binding(Ti) is a vector of length n, holding the binding labels for the target TF in tissue i, and RandomForest(., .) generates the RF model trained on the features and labels provided by the first and second arguments respectively. An example of the input matrix Ti and the response vector Binding(Ti) is shown in Figure 2b. In the second step, to focus only on essential regulators (c.f. Figure 3a), we shrink the feature space to the union of the top 20 regulators taken over all tissue and TF specific RF classifiers, Ti′, by ranking the predictors according to their Gini index (Figure 2c):\n\n\n\nwhere TopFeatures(RFj) denotes the top 20 features of RFj and Subset(., .) generates the reduced feature matrix based on the union of the top TFs. In the following, we refer to a training data set comprised of only one tissue as a single tissue case and to a training data set composed of multiple tissues as a multi tissue case. Considering the single tissue case, we train an RF model, RFi′, on the reduced feature space and use this as the final model for the respective target TF:\n\n\n\nIn the multi-tissue scenario, we retrain tissue-specific RF models on the reduced feature space and apply them across all available training tissues:\n\n\n\nwhere Prediction (RFi′, Ti′) returns the predictions made by RFi′ when applied on the Ti′. Their predictions are combined in a new feature matrix that is used as input to train an ensemble RF, RFE. Note that the input matrix contains predictions of all tissue-specific RF models on all available training tissues (Figure 2d):\n\n\n\nBy design, the ensemble model incorporates the tissue-specific RF classifiers in a non-linear way to better generalize across all provided training tissues. An example matrix that is used to obtain predictions from an ensemble RF is shown in Figure 2e.\n\na) An overview of model training for a distinct transcription factor, TF, with multiple training tissues. Using the full feature matrices T1, T2, T3, depicted in (b), TF and tissue-specific random forest (RF) classifiers are trained. From those RF classifiers (RF1, RF2, RF3), we determine the union of the top 20 features from each RF. In this example, the union of top TFs is comprised of 24 TFs. Next, we design reduced tissue-specific feature matrices T1′,T2′,T3′, as shown in (c) based on the union of the top TF features. Subsequently, tissue-specific RF classifiers (RF1′,RF2′,RF3′) are trained on these reduced feature sets. The tissue-specific RF classifiers are applied to all training tissues and their predictions are aggregated to form the feature matrix TE′, visualized in (d), which is used to train an ensemble model (RFE). The ensemble RF is used to make predictions on unseen data TP′ (e). Note that the column Tissue in d) is not included in the model but only shown here for illustration purposes. The feature matrices shown represent feature setup (1) using DNase1 Hypersensitive (DHS) sites.\n\na) Classification error for the Bound and Unbound classes for different sets of features: considering all features, the top 10, and the top 20 features. One can see that the difference in model performance between the top 20 and all feature cases is only marginal. b) Comparison of the out of bag (OOB) error between ensemble models and tissue-specific random forest (RF) classifiers. Especially in the Unbound case, the ensemble models show superior performance compared to the tissue-specific RF classifiers. c) Misclassification rate computed on unseen test data for ensemble and tissue-specific RF classifiers. As in b) we see that the ensemble models generally outperform the tissue-specific ones. Note that the scale of the y-axis is different for the Bound and Unbound classes in (a) and (b).\n\nWe used two different ways to assess model performance: (1) While fitting the RF classifiers, we measure the out-of-bag error (OOB), which is defined as the mean prediction error for each training sample i using trees that were not trained on sample i. The OOB error is computed separately for the Bound and Unbound classes:\n\n\n\nwhere TP denotes the sites correctly predicted as bound, TN denotes the sites correctly predicted as unbound, FP and FN represent sites incorrectly predicted as bound and unbound, respectively. Note that, because we use balanced data for training the RF classifiers, the OOB is computed on a balanced data set.\n\nAdditionally, we compute (2) the misclassification rate for the Bound and Unbound cases on a subset of the test data that was used by the challenge organizers. The test data is composed of three hold-out chromosomes which have not been used for training: chr1, chr8 and chr21. Additionally, TF binding is predicted on an unseen tissue, i.e. a tissue that was not used for training. An overview of the test data is provided in Table 2. Note that, in contrast to the training data, the test data is not balanced, i.e. the Unbound class is larger than the Bound class. Therefore, to avoid misinterpretation of model performance, it is essential to compute the error for both classes separately.\n\nWe obtained a customized protein-protein-interaction (PPI) probability matrix R as described previously26, which is derived from a random walk analysis on a protein-protein-association network based on STRING27 (version 9.05). An entry Ri,j represents the probability that protein i interacts with protein j. Note that the probability Ri,j is not symmetric by construction, i.e. Ri,j ≠ Rj,i. To generate a score describing how likely it is that a subset of proteins P contained in R interact with a distinct TF t, guided by the feature importances the RF models provide, we define the PPI score St,P as\n\n\n\nwhere GI (p) denotes the Gini index values of p obtained from the RF model corresponding to t. Thus, the smaller the value of St,P the more likely it is that the regulators in P interact with TF t.\n\n\nResults\n\nIn this section, we first show that shrinking the feature space to those TFs essential for training does not affect model accuracy. Next, we demonstrate the benefits of the ensemble learning and how its accuracy is depending on the number of training tissues. We further investigate the top selected TFs by the RF models and find known interaction partners that possess high PPI scores. Finally, we compare the two feature design schemes, described in the Methods section, and explore their influences on model performance. If not stated otherwise, all figures presented in the following are based on annotation setup (1), including DHSs.\n\nBecause having a sparse feature space simplifies model interpretation, we reduce the feature space to contain only a few essential features. As explained above, we determined sets of top features using the Gini index, resulting in TF and tissue-specific sets containing either the top 10 or top 20 features. As shown in Figure 3a the difference in OOB error between the feature set comprised of the top 20 features and the full feature space is only marginal, whereas the difference is increasing when only the top 10 features are considered. Therefore, we decided to use a reduced feature space that consists of the top 20 features per model. The results indicate that the most important feature across all TFs is the DNase1-seq signal within the DHSs for feature setup (1). Similarly, in feature setup (2), the DNase1-seq signal within the bins is found to be more important than the TF features.\n\nAccording to the OOB error shown in Figure 3b, the ensemble RF classifiers outperform the tissue-specific models in all cases for both Bound and Unbound classes, thus emphasizing on the improved capability of the ensemble model to generalize across tissues. Additionally, we computed the misclassification rate on all test tissues which are linked to multiple training tissues (Figure 3c). Again, we notice that the ensemble RF classifiers outperform the tissue-specific classifiers by several orders of magnitude in all Unbound instances and in most Bound cases. Overall, these results suggest that ensemble learning is a promising approach to deal with the tissue-specificity of TF binding.\n\nAlthough the results in Figure 3b and 3c suggest that the ensemble methods perform well, it remains unclear what influence the number of training tissues would have on the performance of an RF. To elucidate this, we performed permutation experiments learning multiple RF models using all possible combinations of training tissues that are available for a distinct TF. As this is a computationally demanding task, we performed it for only three, arbitrarily selected, TFs: MAX, TEAD4, and E2F6. Figure 4a illustrates that the OOB error declines when the number of training tissues increases. Hence, we conclude that the ability of an ensemble RF to generalize across tissues improves with larger number of training tissues.\n\na) Relation of the OOB error for three TFs (E2F6, MAX, and TEAD4) to the number of tissues used for training. The OOB reduces if more tissues are included in the ensemble learning. Red dots represent the mean classification error across all tissue-specific classifiers. Individual models are represented by the black points. b) Comparison between true ensemble models for E2F6, MAX, and TEAD4 and RF classifiers trained on pooled data sets comprised of training data for all available tissues. The ensemble models perform better than the models based on aggregated data.\n\nHowever, it remains to be shown whether the improved accuracy obtained from the ensemble RF classifiers was in fact because of the ensemble learning. To test this, we designed another learning setup in which all tissue-specific data sets were aggregated into one. In other words, we pooled the training data for one TF across all available tissues into one data set. We then used this pooled data set to train a new RF model. As depicted in Figure 4b the true ensemble models perform considerably better than the models learned on the pooled training data. This shows that the ensemble technique is better suited to capture tissue-specific information than simple data aggregation.\n\nAs stated before, we hypothesized that the top predictors selected by the RF classifiers represent regulators that exist either in protein complexes with the target TF via direct or indirect binding, or bind directly to DNA in close proximity to the target TF. To investigate this hypothesis, we computed a PPI score st,P (see Methods) for the selected predictors P per TF t and compared it against scores computed for randomly sampled sets of TFs (based on 100 randomly drawn TF subsets). The PPI score st,P for TF t is small, if t is likely to interact with the factors included in the selected predictor set P. In contrast, the score is high if t is not likely to be interacting with the factors in P. As shown in Figure 5a, except for three TFs (MAX, TAF1, ZNF143), the PPI score of the TFs selected by the RF is better (i.e. smaller) than the scores for the randomly selected set. This indicates that the RF classifiers select features representing regulators that are more likely to be interacting with the target TF, either directly or with indirect contacts.\n\na) Log transformed PPI scores computed for a set of TFs. In the Random case, we show the mean PPI score across 100 random draws and its standard deviation. The smaller the PPI score the better. Only for three TFs (MAX, TAF1, ZNF143), the randomly sampled PPI score is better than or equal to the score derived for the TFs selected by the RF classifiers. b) PPI network obtained from STRING centered around the TF MAFK, highlighting proteins that interact with MAFK with high confidence. Proteins colored in green were identified as important features in the RF classifiers, proteins shown in gray could not be retrieved by our model, because they are DNA-binding proteins, or we do not have a PWM for them in our set. Regulators shown in red could have been detected by the RF, but were not included in the top set of regulators.\n\nFigure 5b provides an example of a PPI network focused on the TF MAFK. The network was obtained from the STRING database27, using the settings highest confidence and no more than 10 interactors. The top features selected by the RF classifiers contain all known regulatory proteins in this network, except for NFE2L2, shown in red. Among these TFs are MAFK itself, MAFF, MAFG and NFE2 (highlighted in green). The strong interactions among the small MAF proteins28 as well as the dimerization of those with NFE229 have been reported in the literature before.\n\nInteraction partners shown in gray can not be identified by our approach as either these are proteins without regulatory functions or we do not have a PWM available for them.\n\nIn the conference round of the challenge, we were using feature setup (1), which is based on DNase1 Hypersensitive Sites (DHSs), while in the final round, we switched to design (2), which is purely based on bins. This transition had a strong effect on our performance assessed by the challenge organizers. While we improved the recall of our predictions by switching from (1) to (2), the precision decreased. In Figure 6, we show the misclassification rates for the Bound and Unbound classes depending on the feature designs. The performance is assessed and shown on test data. The bin based models (2) outperform the peak based models in the Bound case, whereas the peak based models show superior performance in the Unbound case. At the same time, bin based models perform poorly in the Unbound case, which is probably driven by the strong dependence of the RF classifiers on the DNase1-seq signal. In contrast to that, models based on DHSs perform well in the Unbound case, because the search space for TFBSs is limited to only DHSs. This increases the precision of the predictions, but at the same time lowers the recall, which is reflected by the high misclassification rate in the Bound case.\n\nThe bin based model outperforms the peak based model in predicting bound labels, while in the unbound case the model based on DHSs is better than the bin based model.\n\n\nDiscussion and Conclusion\n\nHere, we introduced an RF based ensemble learning approach to predict TFBS in vivo. In this article, we did not compare our approach to competitors in the challenge, as this is done in the main challenge paper. Here, we show the benefits of ensemble learning in a multi-tissue setting and that modeling cofactors is beneficial for the classification.\n\nWe show on both test and training data that the ensemble strategy is able to generalize better across tissues, than models trained on only a single tissue (Figure 3). Also the accuracy of the ensemble classifiers increases with an increasing number of available training tissues (Figure 4a). We also illustrate that just using all available training data to learn one RF does not provide as accurate results as an ensemble model (Figure 4b). In this study, we decided to use RF classifiers, because they lead to accurate classification results using non-linear predictions in a reasonable time. Alternative classification approaches, such as logistic regression, or support-vector-machines could have been used too.\n\nRF classifiers have also been proposed recently, independent from the challenge11, as an adequate method to predict TF binding. Although the authors of11 perform cross cell-type predictions, i.e. they predict TF binding in a tissue where the RF was not trained on, they do not use ensemble models as proposed here. However, they did show that it is beneficial for the predictions of a distinct target TF to consider further TFs as predictors, in addition to the target TF itself. This is in agreement with our findings. As shown in Figure 3a, a small subset of features is sufficient to reach similar classification performance as the full feature space. We found that most of these selected TFs are known interaction partners of the target TF, see Figure 5. This is also supported by a recent study illustrating that most TFs bind in dense clusters around genes suggesting a widespread interaction among them30.\n\nOnly for three TFs, we could not find that the predicted TFs lead to a better PPI score than a randomly chosen set. We note that for two of those three, TAF1 and MAX, the performance of the ensemble RF classifiers improved only marginally, or not at all, compared to the tissue-specific classifiers. This suggests that our model does not account for the true interaction partners of those TFs. Indeed, an inspection of the STRING database for TAF1 revealed that only TAF1 itself and TBP are among the top 20 regulators, which are included in our PWM collection. For the remaining interaction partners, mostly TFs of the TAF family, no binding motif is available in the public repositories, thus they are not included in our PWM collection and can therefore not be used by the RF classifiers. Similarly, for MAX, only 5 out of 20 high confidence interaction partners are included in our PWM collection. Specifically, no PWM is available for 6 TFs interacting with MAX, while the remaining interacting proteins are not categorized as TFs. Overall, our approach benefits from data availability (Figure 4a). If there are only a few TFs available in our PWM collection, it will be harder to model the co-factor binding behavior of a TF across tissues adequately. Also, the more diverse the co-factor landscape of a TF is between the tissues, the harder it will be to learn a general model. Another crucial aspect with respect to that is the quality of the PWM. During the challenge, we realized that the selection of PWMs is crucial for model performance and it is required to compare PWMs obtained from different sources to make sure that one uses the one with highest information content. Nevertheless, instead of using a more recent method to model TF-motifs, we stick to the use of PWMs because they are (1) the most common way to describe the sequence specificity of TFs (2) they are available for a large number of TFs, and (3) they can be interpreted easily.\n\nSwitching the feature design for the RF classifiers from (1) DHS-based to (2) bin-based showed that DHS sites are inadmissible to reduce the false positive rate (Figure 6) of TFBS predictions. Using only bins, without DHS information, we could improve the recall of TFBS predictions, but only at the cost of poor precision at the same time. The explanation for this behavior is a difference in size of the genomic search space between both feature setups. The bin based models have a low misclassification rate in the Bound case, because they do consider the whole genome without neglecting any sites beforehand, thus improving recall. However, our observations suggest that considering only the raw signal does not sufficiently correct for false positive sites, as opposed to use DHSs, which yield an improved misclassification rate in the Unbound case compared to the raw signal.\n\nIn general, both training and evaluating TFBS prediction methods is challenging due to the class imbalance, i.e. there are many more Unbound (negative) than Bound(positive) binding sites in the genome. This requires both (a) training approaches that avoid over-fitting for one of the two classes and (b) evaluation strategies accounting for this issue. Here, we show misclassification rates separately for both positive and negative classes to avoid a bias caused by the dominant Unbound case.\n\nWe note that our current investigation is not meant to construct a genome-wide classifier in which the unbound case is the most abundant. To achieve that, the highly unbalanced training data situation would need to be taken into account, for instance in the loss function of the classifier. Aside from the technical aspects, we show that modeling cofactors is helpful to predict TFBS and that ensemble learning is a promising technique to generalize information across tissues.\n\n\nData availability\n\nThe raw data used in this study is available online at Synapse: https://www.synapse.org/#!Synapse:syn6112317.\n\n\nSoftware availability\n\nCode generated as part of this analysis is available on GitHub: https://github.com/SchulzLab/TFAnalysis\n\nArchived code as time of publication: http://doi.org/10.5281/zenodo.140969731\n\nLicense: MIT", "appendix": "Grant information\n\nThis work was supported by the Cluster of Excellence on Multimodal Computing and Interaction (DFG) [EXC248].\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe thank everyone involved in organizing the ENCODE-DREAM in vivo Transcription Factor binding site prediction challenge and are grateful for the opportunity to share this article. The PPI scoring matrix used in this study was kindly provided by Sebastian Köhler.\n\n\nReferences\n\nVaquerizas JM, Kummerfeld SK, Teichmann SA, et al.: A census of human transcription factors: function, expression and evolution. Nat Rev Genet. 2009; 10(4): 252–263. PubMed Abstract \| Publisher Full Text\n\nNatarajan A, Yardimci GG, Sheffield NC, et al.: Predicting cell-type-specific gene expression from regions of open chromatin. Genome Res. 2012; 22(9): 1711–1722. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMathelier A, Fornes O, Arenillas DJ, et al.: JASPAR 2016: a major expansion and update of the open-access database of transcription factor binding profiles. Nucleic Acids Res. 2016; 44(D1): D110–115. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPique-Regi R, Degner JF, Pai AA, et al.: Accurate inference of transcription factor binding from DNA sequence and chromatin accessibility data. Genome Res. 2011; 21(3): 447–455. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLuo K, Hartemink AJ: Using DNase digestion data to accurately identify transcription factor binding sites. Pac Symp Biocomput. 2013; 80–91. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGusmao EG, Dieterich C, Zenke M, et al.: Detection of active transcription factor binding sites with the combination of DNase hypersensitivity and histone modifications. Bioinformatics. 2014; 30(22): 3143–3151. PubMed Abstract \| Publisher Full Text\n\nKähärä J, Lähdesmäki H: BinDNase: a discriminatory approach for transcription factor binding prediction using DNase I hypersensitivity data. Bioinformatics. 2015; 31(17): 2852–2859. PubMed Abstract \| Publisher Full Text\n\nYardımcı GG, Frank CL, Crawford GE, et al.: Explicit DNase sequence bias modeling enables high-resolution transcription factor footprint detection. Nucleic Acids Res. 2014; 42(19): 11865–11878. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCuellar-Partida G, Buske FA, McLeay RC, et al.: Epigenetic priors for identifying active transcription factor binding sites. Bioinformatics. 2012; 28(1): 56–62. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nO’Connor TR, Bailey TL: Creating and validating cis-regulatory maps of tissue-specific gene expression regulation. Nucleic Acids Res. 2014; 42(17): 11000–11010. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiu S, Zibetti C, Wan J, et al.: Assessing the model transferability for prediction of transcription factor binding sites based on chromatin accessibility. BMC Bioinformatics. 2017; 18(1): 355. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJayaram N, Usvyat D, R Martin AC: Evaluating tools for transcription factor binding site prediction. BMC Bioinformatics. 2016. PubMed Abstract \| Publisher Full Text\n\nKeilwagen J, Grau J: Varying levels of complexity in transcription factor binding motifs. Nucleic Acids Res. 2015; 43(18): e119. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAlipanahi B, Delong A, Weirauch MT, et al.: Predicting the sequence specificities of DNA- and RNA-binding proteins by deep learning. Nat Biotechnol. 2015; 33(8): 831–838. PubMed Abstract \| Publisher Full Text\n\nENCODE-DREAM in vivo transcritpion factor binding site prediction challenge. 2017; Accessed: 2018-02-03. Reference Source\n\nWaardenberg AJ, Homan B, Mohamed S, et al.: Prediction and validation of protein-protein interactors from genome-wide DNA-binding data using a knowledge-based machine-learning approach. Open Biol. 2016; 6(9): pii: 160183. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRoider HG, Kanhere A, Manke T, et al.: Predicting transcription factor affinities to DNA from a biophysical model. Bioinformatics. 2007; 23(2): 134–141. PubMed Abstract \| Publisher Full Text\n\nIbrahim MM, Lacadie SA, Ohler U: JAMM: a peak finder for joint analysis of NGS replicates. Bioinformatics. 2015; 31(1): 48–55. PubMed Abstract \| Publisher Full Text\n\nGrant CE, Bailey TL, Noble WS: Fimo: scanning for occurrences of a given motif. Bioinformatics. 2011; 27(7): 1017–1018. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTanay A: Extensive low-affinity transcriptional interactions in the yeast genome. Genome Res. 2006; 16(8): 962–972. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCrocker J, Abe N, Rinaldi L, et al.: Low affinity binding site clusters confer hox specificity and regulatory robustness. Cell. 2015; 160(1–2): 191–203. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHume MA, Barrera LA, Gisselbrecht SS, et al.: UniPROBE, update 2015: new tools and content for the online database of protein-binding microarray data on protein-DNA interactions. Nucleic Acids Res. 2015; 43(Database issue): D117–122. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKulakovskiy IV, Vorontsov IE, Yevshin IS, et al.: HOCOMOCO: expansion and enhancement of the collection of transcription factor binding sites models. Nucleic Acids Res. 2016; 44(D1): D116–125. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nQuinlan AR, Hall IM: BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics. 2010; 26(6): 841–842. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiaw A, Wiener M: Classification and regression by randomforest. R News. 2002; 2(3): 18–22. Reference Source\n\nKöhler S, Bauer S, Horn D, et al.: Walking the interactome for prioritization of candidate disease genes. Am J Hum Genet. 2008; 82(4): 949–958. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSzklarczyk D, Morris JH, Cook H, et al.: The STRING database in 2017: quality-controlled protein-protein association networks, made broadly accessible. Nucleic Acids Res. 2017; 45(D1): D362–D368. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKannan MB, Solovieva V, Blank V: The small MAF transcription factors MAFF, MAFG and MAFK: current knowledge and perspectives. Biochim Biophys Acta. 2012; 1823(10): 1841–1846. PubMed Abstract \| Publisher Full Text\n\nIgarashi K, Kataoka K, Itoh K, et al.: Regulation of transcription by dimerization of erythroid factor NF-E2 p45 with small Maf proteins. Nature. 1994; 367(6463): 568–572. PubMed Abstract \| Publisher Full Text\n\nYan J, Enge M, Whitington T, et al.: Transcription factor binding in human cells occurs in dense clusters formed around cohesin anchor sites. Cell. 2013; 154(4): 801–813. PubMed Abstract \| Publisher Full Text\n\nSchulzLab, Schmidt F: Florian411/TFAnalysis: Release for F1000 article (Version 1.0). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1409697" }	[ { "id": "39062", "date": "26 Oct 2018", "name": "Jan Grau", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nTranscriptional regulation by transcription factors (TFs) is one of the fundamental steps of gene regulation. Hence, knowing the genome-wide binding regions of a TF is of great interest. Experimentally, those could be determined by ChIP-seq, which, however, is time-consuming and labor-intensive. Hence, computational prediction of cell type-specific, in-vivo transcription factor binding is highly demanded. In their manuscript \"Predicting transcription factor binding using ensemble random forest models\", Ardakani, Schmidt and Schulz present a novel method for this purpose, which is based on PWMs describing TF sequence preference, and DNase-seq data capturing chromatin accessibility. This method combines i) learning random forest (RF) classifiers on feature matrices for individual cell types, ii) shrinking feature sets, and iii) learning ensemble classifiers across cell types. The authors illustrate that within their method, peak-based DNase features seem to be favorable compared with bin-based aggregation of DNase-seq coverage. Furthermore, they demonstrate that the ensemble classifier indeed yields an overall improved performance compared with cell type-specific RFs. As this is a companion paper to the main publication describing the results of the ENCODE-DREAM challenge, I consider a direct comparison to other approaches dispensable in this case.\nIn general, most of the methods are well described and conclusions are supported by the data. However, I have a few major and several minor comments regarding choices made by the authors (especially with regard to performance assessment) and the presentation of specific details of methods and results, as detailed in the following.\n\nMajor comments:\n1. In sub-section \"Data\" of the Methods section, the authors state that they \"focus on these 12 TFs in the scope of this article\". However, this is contradicted by the list provided in Table 2 listing only 8 TFs. Results for the same 8 TFs are also shown in Fig. 6, whereas several of the remaining figures (Fig 3a/b, Fig 5) present results for a larger set of TFs, i.e., for TFs not listed in sub-section \"Data\".\n2. The third paragraph of sub-section \"Data preprocessing and feature generation\" of the Methods section is lacking details. How exactly are \"tissue-specific DHSs\" called using JAMM? What have been the inputs and input formats? Which peaks are merged and why?\n3. Results with regard to feature shrinkage (Fig. 3a) are only shown for OOB Misclassification. As I could imagine over-fitting effects to specifics of the training cell types, I considered an evaluation on the test data highly informative. For instance, I would imagine that we see a decrease in OOB performance when shrinking features to the top 20, whereas on the test data this model achieves a better generalization and, hence, misclassification rate.\n4. The authors chose to use misclassification separated by classes, which could also be described as false negative rate and false positive rate, as performance measure for the whole manuscript. For several reasons, I would consider curve-based measures, especially the (area under the) precision-recall curve the more appropriate measure for this application but also in the context of the ENCODE-DREAM challenge. First, we face a highly imbalanced classification problem, and the precision-recall curve has been shown to be highly informative in this case1. Second, the areas under the ROC curve and precision-recall curve have also been used for performance assessment in the ENCODE-DREAM challenge and choosing the same performance measure in this paper would foster comparison of results to those of the challenge (especially since both use the same test data). Third, in the discussion of Fig. 6, the authors mention that one choice of DNase data works better for bound regions, while the other works better for unbound regions. Here, we face the typical trade-off between sensitivity and specificity (or false negative rate and false positive rate), where we are unable to decide for one option based on specific, contradictory combinations of the two measures. In the ROC curve, basically (1 - FN/(TP+FN)) would be plotted against FP/(TN+FP), so we would get a broader impression of classifier performance, including the specific points on the curve chosen by the authors. For these reasons, the area under the ROC curve and the area under the precision-recall curve should be included as performance measures into this study. As the authors illustrate in Fig. 2d, RF classifiers already output continuous scores that could be used for computing these curves. Technically, curves and AUC values could be computed, e.g., using the R packages PRROC or precrec.\n5. In sub-section \"Ensemble learning improves model accuracy\" of the Results section, I agree with the authors that the ensemble classifier performs better than the individual RFs. However, currently it remains unclear if this can really be attributed to \"ensemble learning\" or just to averaging effects. Hence, I would suggest to include a simple averaging over the predictions of individual RFs (those, for which the predictions are also input of RF_E) as a simple baseline model (in addition to the single RF learned on the pooled data). In addition, for MAX, the authors might also include results for the test data in addition to what is shown in Figure 4.\n\nMinor comments:\n6. In the Introduction, second paragraph, the authors state that \"Most of these methods are based on position weight matrices (PWMs) describing the sequence preference of TFs,\" giving a reference to the publication of the 2016 update of the Jaspar database. While Jaspar indeed provides PWM models, I do not consider this an appropriate reference for the definition of PWMs in general. Specifically, I would suggest to cite the seminal works of Berg & von Hippel 2 and of Stormo3 instead.\n7. In the Introduction, second paragraph, the authors state \"PWMs indicate [...] which nucleotide is most likely to occur\". From my perspective, this description is not fully accurate. The most likely nucleotide is also represented by consensus sequences. PWMs give a specific weight (or log-probability,...) for each of the nucleotides and not only for the most likely one.\n8. I appreciate that the authors reference our work regarding dependency models (Slim models) in the second paragraph of the introduction. However, there are several other approaches for modeling dependencies in TF binding sites. I would encourage the authors to broaden the scope of their references by including, e.g.4-5.\n9. In the third paragraph of the introduction, the authors refer to \"the main ENCODE-DREAM Challenge paper\". I am aware that this paper has not yet been published, but encourage the authors to update their publication including a reference to that paper when available.\n10. In the second paragraph of sub-section \"Data preprocessing and feature generation\" of the Methods section, it is mentioned that TF binding affinities are computed for 557 distinct TFs. After reading the complete paper, I understood (hopefully correctly) that all 557 TFs are used for all RFs (before shrinking the feature space) regardless of the training TF. If my understanding is correct, the authors might consider to include an explicit statement about this fact already at this stage of the manuscript.\n11. In the first paragraph of sub-section \"Ensemble random forest classifier\" of the Methods section, the authors state that \"the balance between the bound and unbound classes is maintained to avoid over-fitting\". For me, it remains unclear how exactly this helps to avoid over-fitting. For my understanding, over-fitting typically refers to an over-adaption to specifics of the training data, which do not generalize well to other data sets, leading to a poor performance on unseen (test) data. However, the class imbalance is inherent to the problem and should be (roughly) the same for training and test cell types. Please clarify.\n12. In the first paragraph of sub-section \"Ensemble random forest classifier\" of the Methods section, right before the second formula, the shrunken feature space is described to be the union of top 20 regulators. However, later in the Results section, the authors also consider a case where features per RF are restricted to the top 10 ones (Fig 3a). Hence, I would suggest a generic description, here.\n13. The third formula of sub-section \"Ensemble random forest classifier\" of the Methods section refers to an index i, where (for my understanding), according to the previous definition, i should be in {1}, in this case. If that is indeed the case, I would suggest to replace \"i\" by \"1\" in the formula and explicitly state that this is the only index i can be.\n14. The fourth formula of sub-section \"Ensemble random forest classifier\" of the Methods section is partly broken. Specifically, the element sign refers to the set of indexes, which does not seem reasonable to me. I rather think this should refer to the matrix resulting from prediction(RF_i',T_i') Please fix.\n15. In Figure 2 (b), (c) and (e), the labels in the table cells are hardly legible in printout. Either increase the thickness of letters or chose a different color.\n16. For Figure 2e, it remains unclear from the caption what is shown. It seems to be the input matrix derived from test data, in analogy to the training matrices shown in Figure 2b? Is this the input of each RF? Of RF' (as features might have been shrunken)? Or of RF_E?\n17. The fifth formula of sub-section \"Ensemble random forest classifier\" of the Methods might profit from a bit of additional explanation. Specifically, it took me a while to understand (if I'm right) that for T_E', the outputs of all individual RFs are concatenated row-wise, while \"Binding(T_E')\" denotes the concatenation of training labels.\n18. In the first paragraph of sub-section \"Performance assessment\" of the Methods section, I wondered what the index \"i\" refers to. Is this the same index i as before (i.e., an index for the training cell types)? If not, what exactly is \"sample i\"?\n19. In sub-section \"Protein-protein-interaction score\" of the Methods section, I would have appreciated a bit more motivation before describing the method itself.\n20. In sub-section \"Reducing the feature space to a small subset [...]\" of the Results section, I would not fully agree with the authors that the difference in error between the full model and the model based on top 20 features is \"marginal\". I would even assume that a statistical test of the difference between the data behind the two boxplots in Fig. 3a would be significant.\n21. In sub-section \"Reducing the feature space to a small subset [...]\" of the Results section, I did not find the last two sentences (regarding importance of DNase-based features) to be supported by the data shown in the manuscript.\n22. In section \"Data availability\", the authors provide a link to the synapse page of the ENCODE-DREAM challenge. However, the data are accessible only after registration and signing a data usage policy.\n23. Typos & Grammar: - first paragraph of \"Data preprocessing and feature generation\": \"down sampled\" should be \"down-sampled\" - second paragraph of \"Data preprocessing and feature generation\": \"the course of challenge\" should be \"the course of the challenge\" - third paragraph of \"Data preprocessing and feature generation\": \"data is intersected\" should be \"data are intersected\" - 7th paragraph of \"Discussion and conclusions\": \"Bound(positive)\" should be \"Bound (positive)\" - Reference 15: \"transcritpion\" should be \"transcription\"\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4783", "date": "08 Aug 2019", "name": "Fatemeh Behjati Ardakani", "role": "Author Response", "response": "1. In sub-section \"Data\" of the Methods section, the authors state that they \"focus on these 12 TFs in the scope of this article\". However, this is contradicted by the list provided in Table 2 listing only 8 TFs. Results for the same 8 TFs are also shown in Fig. 6, whereas several of the remaining figures (Fig 3a/b, Fig 5) present results for a larger set of TFs, i.e., for TFs not listed in sub-section \"Data\". We thank the reviewer for spotting this. There was indeed an error in Table 1 and some TFs were missing. We have corrected Table 1 to list all TFs considered in Figures 3 and 5. In Fig.6, as well as in Fig.3c, we show results on test data from the challenge, therefore there are fewer TFs than in the remaining Figures. As we only look at the multi-tissue cases, for which there are more than one training tissue per TF available, we use only a subset of the available challenge data. ===================================================================================== 2. The third paragraph of sub-section \"Data preprocessing and feature generation\" of the Methods section is lacking details. How exactly are \"tissue-specific DHSs\" called using JAMM? What have been the inputs and input formats? Which peaks are merged and why? We have clarified these points in the main text. We converted the provided DNase1-seq bam files to bed files using the bedtools bamtobed command. For each bed file, peaks are computed separately using JAMM’s standard parameters and the –f 1 option. The individual DHS files obtained for one TF are aggregated using the bedtools merge command. We decided to take a less conservative approach and merge all peaks identified in individual replicates per TF to ensure that we do not miss an accessible site, all be it this may introduce false positives. ===================================================================================== 3. Results with regard to feature shrinkage (Fig. 3a) are only shown for OOB Misclassification. As I could imagine over-fitting effects to specifics of the training cell types, I considered an evaluation on the test data highly informative. For instance, I would imagine that we see a decrease in OOB performance when shrinking features to the top 20, whereas on the test data this model achieves a better generalization and, hence, misclassification rate. We appreciate the suggestions and performed the same experiment as shown in Figure 3a using the challenge’s test data (Supplementary Figure 1). As expected, we find a slight decrease in terms of OOB performance for the top10 and top20 models compared to all features, whereas on test data we see that both the top10 and top20 models perform slightly better than models considering all features. However, we note that the differences are not significant. ===================================================================================== 4. The authors chose to use misclassification separated by classes, which could also be described as false negative rate and false positive rate, as performance measure for the whole manuscript. We have mentioned these more established names in the main manuscript. However, we decided to stick to the already used nomenclature, as we believe that it is more comprehensible in the context of the TF binding predictions. For several reasons, I would consider curve-based measures, especially the (area under the) precision-recall curve the more appropriate measure for this application but also in the context of the ENCODE-DREAM challenge. First, we face a highly imbalanced classification problem, and the precision-recall curve has been shown to be highly informative in this case1. Second, the areas under the ROC curve and precision-recall curve have also been used for performance assessment in the ENCODE-DREAM challenge and choosing the same performance measure in this paper would foster comparison of results to those of the challenge (especially since both use the same test data). Third, in the discussion of Fig. 6, the authors mention that one choice of DNase data works better for bound regions, while the other works better for unbound regions. Here, we face the typical trade-off between sensitivity and specificity (or false negative rate and false positive rate), where we are unable to decide for one option based on specific, contradictory combinations of the two measures. In the ROC curve, basically (1 - FN/(TP+FN)) would be plotted against FP/(TN+FP), so we would get a broader impression of classifier performance, including the specific points on the curve chosen by the authors. For these reasons, the area under the ROC curve and the area under the precision-recall curve should be included as performance measures into this study. As the authors illustrate in Fig. 2d, RF classifiers already output continuous scores that could be used for computing these curves. Technically, curves and AUC values could be computed, e.g., using the R packages PRROC or precrec. We agree with the reviewer that curve based scores like PR and ROC are better to assess the performance of our models. As suggested, we used the PRROC package to compute AUC values for PR and ROC curves and use these measures throughout the article. It is worth noting that the PR values deliver a more realistic impression on model performance than ROC or the misclassification rate on the highly unbalanced test data sets, which are enriched with negative cases, i.e. unbound sites. We moved the previous figures using the misclassification rate to the Supplement. ==================================================================================== 5. In sub-section \"Ensemble learning improves model accuracy\" of the Results section, I agree with the authors that the ensemble classifier performs better than the individual RFs. However, currently it remains unclear if this can really be attributed to \"ensemble learning\" or just to averaging effects. Hence, I would suggest to include a simple averaging over the predictions of individual RFs (those, for which the predictions are also input of RF_E) as a simple baseline model (in addition to the single RF learned on the pooled data). We agree with the reviewer's comment, and as suggested, we added another model averaging over the predictions of the tissue specific RFs as a baseline for our ensemble models. As shown in Figure 4b, the averaging leads to a worse performance than simply pooling the information across all samples into one model, indicating that the ensemble step does indeed combine tissue specific information in a more sophisticated way than a simple average. In addition, for MAX, the authors might also include results for the test data in addition to what is shown in Figure 4. In the interest of clarity and homogeneity of the analysis, we refrained from performing the analysis for MAX additionally on test data. Minor comments: 6. In the Introduction, second paragraph, the authors state that \"Most of these methods are based on position weight matrices (PWMs) describing the sequence preference of TFs,\" giving a reference to the publication of the 2016 update of the Jaspar database. While Jaspar indeed provides PWM models, I do not consider this an appropriate reference for the definition of PWMs in general. Specifically, I would suggest to cite the seminal works of Berg & von Hippel 2 and of Stormo 3 instead. We agree with the reviewer and have changed the citation. ==================================================================================== 7. In the Introduction, second paragraph, the authors state \"PWMs indicate [...] which nucleotide is most likely to occur\". From my perspective, this description is not fully accurate. The most likely nucleotide is also represented by consensus sequences. PWMs give a specific weight (or log-probability,...) for each of the nucleotides and not only for the most likely one. This is true. We adapted the wording in the main text to avoid the confusion. ===================================================================================== 8. I appreciate that the authors reference our work regarding dependency models (Slim models) in the second paragraph of the introduction. However, there are several other approaches for modeling dependencies in TF binding sites. I would encourage the authors to broaden the scope of their references by including, e.g.4-5. We appreciate the hint and have included the suggested literature. ===================================================================================== 9. In the third paragraph of the introduction, the authors refer to \"the main ENCODE-DREAM Challenge paper\". I am aware that this paper has not yet been published, but encourage the authors to update their publication including a reference to that paper when available. Yes, we will cite this paper once it is available. Up to the submission of this revised version of our article, the challenge paper has not yet been published. ===================================================================================== 10. In the second paragraph of sub-section \"Data preprocessing and feature generation\" of the Methods section, it is mentioned that TF binding affinities are computed for 557 distinct TFs. After reading the complete paper, I understood (hopefully correctly) that all 557 TFs are used for all RFs (before shrinking the feature space) regardless of the training TF. If my understanding is correct, the authors might consider to include an explicit statement about this fact already at this stage of the manuscript. Yes, the modelling is performed exactly in that way. We have improved the wording to make this more pronounced at the specified position in the main text. ===================================================================================== 11. In the first paragraph of sub-section \"Ensemble random forest classifier\" of the Methods section, the authors state that \"the balance between the bound and unbound classes is maintained to avoid over-fitting\". For me, it remains unclear how exactly this helps to avoid over-fitting. For my understanding, over-fitting typically refers to an over-adaption to specifics of the training data, which do not generalize well to other data sets, leading to a poor performance on unseen (test) data. However, the class imbalance is inherent to the problem and should be (roughly) the same for training and test cell types. Please clarify. Yes, the term “overfitting” might have been inaccurate when class imbalance was considered. We mean the class imbalance effects on training would've been attenuated by keeping the balance between bound an unbound in our training data. The class distribution of the test data, however, would not be a problem given that the models are fairly and reliably trained. ===================================================================================== 12. In the first paragraph of sub-section \"Ensemble random forest classifier\" of the Methods section, right before the second formula, the shrunken feature space is described to be the union of top 20 regulators. However, later in the Results section, the authors also consider a case where features per RF are restricted to the top 10 ones (Fig 3a). Hence, I would suggest a generic description, here. We thank the reviewer to point out this inconsistency. We have replaced the numeric value by a generic description. ===================================================================================== 13. The third formula of sub-section \"Ensemble random forest classifier\" of the Methods section refers to an index i, where (for my understanding), according to the previous definition, i should be in {1}, in this case. If that is indeed the case, I would suggest to replace \"i\" by \"1\" in the formula and explicitly state that this is the only index i can be. The observation is correct. We adapted the text accordingly. ===================================================================================== 14. The fourth formula of sub-section \"Ensemble random forest classifier\" of the Methods section is partly broken. Specifically, the element sign refers to the set of indexes, which does not seem reasonable to me. I rather think this should refer to the matrix resulting from prediction(RF_i',T_i') Please fix. We thank the reviewer for spotting this mistake. We have corrected it. ===================================================================================== 15. In Figure 2 (b), (c) and (e), the labels in the table cells are hardly legible in printout. Either increase the thickness of letters or chose a different color. We have increased the font size. ===================================================================================== 16. For Figure 2e, it remains unclear from the caption what is shown. It seems to be the input matrix derived from test data, in analogy to the training matrices shown in Figure 2b? Is this the input of each RF? Of RF' (as features might have been shrunken)? Or of RF_E? Indeed, in Figure 2e the input matrix for the test instances is shown. The matrix is used as input for the individual classifiers T′1, T′2, T′3, which is the classifiers learned on the reduced feature space. We have improved the figure legend to better address this point. ===================================================================================== 17. The fifth formula of sub-section \"Ensemble random forest classifier\" of the Methods might profit from a bit of additional explanation. Specifically, it took me a while to understand (if I'm right) that for T_E', the outputs of all individual RFs are concatenated row-wise, while \"Binding(T_E')\" denotes the concatenation of training labels. We reformulated the text for better clarity. ===================================================================================== 18. In the first paragraph of sub-section \"Performance assessment\" of the Methods section, I wondered what the index \"i\" refers to. Is this the same index i as before (i.e., an index for the training cell types)? If not, what exactly is \"sample i\"? We have removed the index. It was not required at this point. ===================================================================================== 19. In sub-section \"Protein-protein-interaction score\" of the Methods section, I would have appreciated a bit more motivation before describing the method itself. We have added a sentence for motivation. ===================================================================================== 20. In sub-section \"Reducing the feature space to a small subset [...]\" of the Results section, I would not fully agree with the authors that the difference in error between the full model and the model based on top 20 features is \"marginal\". I would even assume that a statistical test of the difference between the data behind the two boxplots in Fig. 3a would be significant. We performed a statistical test on the difference of PR-AUC and ROC-AUC for both the OOB error as well as the test data (Figure 3a and Sup. Fig1, respectively). The differences are not significant for any of those instances. ===================================================================================== 21. In sub-section \"Reducing the feature space to a small subset [...]\" of the Results section, I did not find the last two sentences (regarding importance of DNase-based features) to be supported by the data shown in the manuscript. We appreciate that the reviewer pointed us to the lack of evidence required for this statement. We have added another Figure (Figure 7) to the main paper illustrating the feature importance of the RFs, which supports the statement made in the section mentioned above. ===================================================================================== 22. In section \"Data availability\", the authors provide a link to the synapse page of the ENCODE-DREAM challenge. However, the data are accessible only after registration and signing a data usage policy. We thank the reviewer for pointing this out to us. We have added it to the main manuscript. ===================================================================================== 23. Typos & Grammar: - first paragraph of \"Data preprocessing and feature generation\": \"down sampled\" should be \"down-sampled\" - second paragraph of \"Data preprocessing and feature generation\": \"the course of challenge\" should be \"the course of the challenge\" - third paragraph of \"Data preprocessing and feature generation\": \"data is intersected\" should be \"data are intersected\" - 7th paragraph of \"Discussion and conclusions\": \"Bound(positive)\" should be \"Bound (positive)\" - Reference 15: \"transcritpion\" should be \"transcription\" We thank the reviewer for spotting the typos, we have corrected them." } ] }, { "id": "42084", "date": "07 Jan 2019", "name": "Gary D Stormo", "expertise": [ "Reviewer Expertise My expertise is in computational and experimental studies of protein-DNA interactions and the regulation of gene expression", "which are relevant to this work." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper reports the results of this group's entry into the ENCODE-DREAM challenge. The task of the challenge was to learn a model for binding of a target TF based on ChIP-seq data and DHS data from different cell types, and to predict binding on held-out data. They focused on a subset of 12 TFs. There were two types of held-out data, three chromosomes from the same cell types as the training data, and also data from different cell types not used in training. Results are reported as classification errors independently for bound and unbound sites.\nThis group did not try to learn a model (such as PWM) for the target TF, rather they used existing PWM models, available in databases, for the target TF as well as for 556 other TFs (so 557 in total; when more than one PWM was available for a TF they used the one with the highest information content). They employed a random forest (RF) approach for learning the model, and they compared variations on how the training was performed.\n\nThere isn't yet a summary publication of the results of the challenge, so at this time we do not know how this approach compares to others. But there are some results reported that are interesting to know regardless of the ranking of this approach.\nOne variation they tested was training using all of the features (a DHS score and all of the PWM scores) versus only subsets, and ranking the features to see which are most important. They found that using only the top 20 features was essentially as good as all of them, whereas the top 10 was not. Not surprisingly, the DHS score is the most important feature. They don't state it, but I assume that the PWM for the target TF is the next most important. Is that right? It is also reassuring that the set of other TFs that rank highest in importance are enriched for TFs previously shown to interact with the target TF, indicating that their models are learning something about the coordinated regulation by multiple TFs.\nThey also compared prediction accuracy on models trained on individual tissue type data, versus a model trained on all of the tissue data merged together, versus an ensemble model obtained from all of the tissue types, with each treated independently. The ensemble models performed significantly better than the others (although I would like to see a separation of results on the different types of test data - see comments below). And the models improve with additional tissue types, although for most TFs the improvement is marginal beyond three.\nComments and suggestions:\n1. Their reporting of results is less informative than it could be. For example, instead of just reporting a classification error for each class (bound and unbound) they could show ROC or PRC curves that indicate those errors for a range of thresholds. Is the reason they don't do that because their program simply returns a binary result, bound/unbound, rather than a probability (or some quantitative score) of being bound? The results as reported highlight the intrinsic tradeoff between false positive and false negative predictions because they vary rather dramatically between different test sets, but don't provide any guidance of how one might balance the two to obtain \"optimal\" predictions (where optimal may depend on the usage).\n2. In Figure 3c they show results on the two types of held-out data, from left out chromosomes from the same tissues as the training data and from data from different tissues. I would like to see those two types of test data reported separately. I can easily imagine that testing on left out chromosomes from the same tissue would provide better predictions, because the same set of additional TFs are utilized within the same tissue, but that on different tissues that might not be the case and that the ensemble approach might be especially useful.\n3. I'm a little confused about the differences in the two training methods shown in Figure 1, and I think some clarification is needed. 1a is clear enough, they are just using genomic regions under DHS peaks (in a given tissue), and the training involves those that are bound by the TF (in that same tissue) and those that are not. But in 1b, is the whole genome binned (and what are bin sizes, I didn't see that stated)? And then is the training on that whole genome, so that the unbound training data enormously larger than the bound data (in fact the vast majority of the genome is not under DHS peaks so its relevance isn't clear). And then when testing the models obtained from the binned training, do they make predictions on the whole genome, or only on the DHS regions? They report that training on binned data was better, but it isn't clear to me is the assessments were the same (such as testing on the whole genome versus under the DHS peaks) which may make a difference.\n4. The word \"inadmissible\" occurs twice, once in the Introduction and once in the Discussion. It doesn't seem to be the right word in either case, in fact based on the context I think it is opposite of what they mean. For example, the first occurrence is \"(TFs) are inadmissible to maintain and establish cellular identity....\". I think \"essential\" or \"required\" are more appropriate.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4784", "date": "08 Aug 2019", "name": "Fatemeh Behjati Ardakani", "role": "Author Response", "response": "1.Their reporting of results is less informative than it could be. For example, instead of just reporting a classification error for each class (bound and unbound) they could show ROC or PRC curves that indicate those errors for a range of thresholds. Is the reason they don't do that because their program simply returns a binary result, bound/unbound, rather than a probability (or some quantitative score) of being bound? We agree with the reviewer that ROC and PR curves are meaningful error measures. We did not choose those initially as we believed that the misclassification rate is a more intuitive measure. Our models do allow us to compute ROC and PR curves. In the revised version of the article, we report the area under the precision recall curve (AU-PR) as well as the area under the receiver operator characteristic curve (AUC-ROC). We have moved the misclassification to the Supplement. The results as reported highlight the intrinsic tradeoff between false positive and false negative predictions because they vary rather dramatically between different test sets, but don't provide any guidance of how one might balance the two to obtain \"optimal\" predictions (where optimal may depend on the usage). We thank the reviewer for pointing out to us that the two proposed setups could be combined. It is a thought that did not occur to us. One option would be to combine the predictions obtained using both feature setups in yet another ensemble RF model. The balancing could be controlled by a customized penalization of the model, such that either Precision, Recall, or both are optimized. We addressed this point in the discussion of our article. ===================================================================================== 2. In Figure 3c they show results on the two types of held-out data, from left out chromosomes from the same tissues as the training data and from data from different tissues. I would like to see those two types of test data reported separately. I can easily imagine that testing on left out chromosomes from the same tissue would provide better predictions, because the same set of additional TFs are utilized within the same tissue, but that on different tissues that might not be the case and that the ensemble approach might be especially useful. We thank the reviewer for this suggestion. Indeed we see that the ensemble model predicting tissue X as well as the classifier trained only on chromosomes of tissue X, perform equally well. In contrast when evaluating the classifiers on other cell types, the ensemble method performs better than any other classifier trained on only one tissue. The results are shown in Supplementary Figure 2. ===================================================================================== 3. I'm a little confused about the differences in the two training methods shown in Figure 1, and I think some clarification is needed. 1a is clear enough, they are just using genomic regions under DHS peaks (in a given tissue), and the training involves those that are bound by the TF (in that same tissue) and those that are not. But in 1b, is the whole genome binned (and what are bin sizes, I didn't see that stated)? And then is the training on that whole genome, so that the unbound training data enormously larger than the bound data (in fact the vast majority of the genome is not under DHS peaks so its relevance isn't clear). And then when testing the models obtained from the binned training, do they make predictions on the whole genome, or only on the DHS regions? They report that training on binned data was better, but it isn't clear to me is the assessments were the same (such as testing on the whole genome versus under the DHS peaks) which may make a difference. We agree with the reviewer that this is a bit unclear without more detailed information on the challenge setup itself. We have added a description on the training, test, and benchmarking data provided by the challenge organizers to the main text. As stated there, the challenge’s objective was to predict TF binding in bins of size 200bp, shifted by 50bp each. Predictions are computed for all bins in chromosomes 1, 8, and 21, the remaining chromosomes are used for training. To train the models, all bound bins in training chromosomes as well as an equal number of randomly sampled unbound bins have been used. The DNase1-seq signal in these bins is what is used in the setup described in Figure 1b. We believed that using the RFs to learn an association between DNase1-seq signal and TF binding might outperform a peak-calling based method, therefore we have pursued this approach as well. The models are assessed on the bin level for both setups. In Setup 1 (Fig1a), any bin not overlapping a DHS is predicted as unbound per default, bins overlapping a DHS are subjected to classification. In Setup 2 (Fig1b) each bin is classified. Thus, the setups can be compared. We have improved the description of Setup 2 (Fig1b) in the main text. ===================================================================================== 4. The word \"inadmissible\" occurs twice, once in the Introduction and once in the Discussion. It doesn't seem to be the right word in either case, in fact based on the context I think it is opposite of what they mean. For example, the first occurrence is \"(TFs) are inadmissible to maintain and establish cellular identity....\". I think \"essential\" or \"required\" are more appropriate. We thank the reviewer for spotting this mistake. We meant to say indispensable." } ] } ]	1	https://f1000research.com/articles/7-1603
https://f1000research.com/articles/8-1566/v1	02 Sep 19	{ "type": "Case Report", "title": "Case Report: The first probable Hong Kong Chinese case of LPIN1-related acute recurrent rhabdomyolysis in a boy with two novel variants", "authors": [ "Sau Wing Yim", "Tina Yee Ching Chan", "Kiran M. Belaramani", "Sze Shun Man", "Felix Chi Kin Wong", "Sammy Pak-Lam Chen", "Hencher Han Chih Lee", "Chloe Miu Mak", "Chor Kwan Ching", "Sau Wing Yim", "Tina Yee Ching Chan", "Kiran M. Belaramani", "Sze Shun Man", "Felix Chi Kin Wong", "Sammy Pak-Lam Chen", "Hencher Han Chih Lee", "Chloe Miu Mak" ], "abstract": "Recurrent rhabdomyolysis is frequently ascribed to fatty acid ß-oxidation defects, mitochondrial respiratory chain disorders and glycogen storage-related diseases. In recent years, autosomal recessive LPIN1 mutations have been identified as a prevailing cause of severe rhabdomyolysis in children in Western countries. We report the first probable Hong Kong Chinese case of recurrent severe rhabdomyolysis in early childhood caused by LPIN1 variants. Compound heterozygous novel variants NM_145693.2(LPIN1):c.[1949_1967dupGTGTCACCACGCAGTACCA]; [2410G>C] (p.[Gly657Cysfs12];[Asp804His]) were detected. The former variant was classified as likely pathogenic while the latter variant was classified as a variant of uncertain significance (VUS) based on the guideline published by the American College of Medical Genetics and Genomics (ACMG) in 2015. Although the genetic findings were inconclusive, the patient’s presentation was compatible with LPIN1-related acute recurrent rhabdomyolysis, and the patient was treated as such. The early recognition, timely diagnosis and management of this condition are important to avoid fatal consequences. To our knowledge, there has been no previous report in the English-language literature of a child with Chinese ethnicity and LPIN1-related acute recurrent rhabdomyolysis (MIM #268200). Functional characterization of the novel variants detected in this study are warranted in future studies.", "keywords": [ "Hong Kong Chinese", "LPIN1", "Rhabdomyolysis", "Novel variants" ], "content": "Introduction\n\nRhabdomyolysis is an uncommon but potentially fatal condition. It results from an acute muscle fiber breakdown resulting in the release of intracellular muscle constituents into the circulation. There are multiple causes for rhabdomyolysis. However, recurrent rhabdomyolysis is frequently attributed to metabolic myopathies. The differential diagnoses include fatty acid ß-oxidation defects (FAOD), mitochondrial respiratory chain disorders and glycogen storage diseases. When initial investigations including biochemical metabolic work-up are unremarkable, muscle biopsies for histological and enzymatic studies are commonly performed. In recent years, autosomal recessive mutations in LPIN1 are being increasingly recognized as an important cause of acute recurrent rhabdomyolysis in childhood in the Western population, after the exclusion of FAOD, mitochondrial respiratory chain disorders and glycogen storage diseases. Lipin 1 (LPIN1), encoded by LPIN1 gene, belongs to the LPIN gene family which also include LPIN2 and LPIN3. Lipin 1 is predominantly expressed in skeletal muscle and adipose tissue1–3. Among lipins 1, 2 and 3, lipin 1 is the only lipin which is expressed significantly in muscles2,4. Lipin 1 is a Mg2+-dependent phosphatidic acid phosphohydrolase (PAP) that catalyzes the dephosphorylation of phosphatidic acid to yield diacylglycerol and inorganic phosphate5. It has also been found to exert transcriptional co-regulator activity6,7. The human LPIN1 gene is mapped to chromosome 2p25.1 and contains 20 exons. Apart from acute recurrent rhabdomyolysis, LPIN1 variants have also been found in adults with statin-induced myopathy, metabolic syndrome and type 2 diabetes mellitus8–11. We hereby report the first probable Hong Kong Chinese case of acute recurrent rhabdomyolysis in a boy with compound heterozygous LPIN1 variants who presented with recurrent rhabdomyolysis since the age of 15 months.\n\n\nCase report\n\nA 15-month-old boy was brought to the hospital complaining of coryzal symptoms for three days, and dyspnea and fever for one day. He was seen by a general practitioner a day prior to the admission and was given an intramuscular injection of an uncertain drug for his illness. After the injection, he complained of passing dark-colored urine twice. Physical examination showed good perfusion and the patient’s blood pressure was 113/65 mmHg. His body weight and height were at the 75th percentile and 50th percentile, respectively, for his age and sex. Respiratory examination revealed bilateral wheezes and the initial impression was acute bronchiolitis. Creatine kinase (CK) was markedly elevated to 127,494 U/L (reference interval: 39–308 U/L). Urine dipstick was positive for hemoglobin. After centrifugation of the urine through a 30 kDa microconcentrator membrane, urine dipstick for hemoglobin remained positive, confirming that the positive urine dipstick result without ultrafiltration was likely due to the presence of myoglobin, which cross-reacts with the peroxidase reaction employed by the urine dipstick for hemoglobin testing. Plasma creatinine and potassium were normal. Liver function tests revealed an elevated alanine aminotransferase (ALT) of 658 U/L (reference interval: 9–25 U/L) on the first day of admission; the value peaked on the second day of admission with a reading of 1,418 U/L. Apart from CK, lactate dehydrogenase (LDH) level was also raised to 9,154 U/L (reference interval: <350 U/L) and it peaked on the second day of admission at a value of 11,409 U/L. Cardiac troponin-I level was normal and electrocardiogram (ECG) did not show any abnormalities. A diagnosis of rhabdomyolysis was made.\n\nIntravenous fluid was given in view of poor appetite and later tapered off as diet was better tolerated. CK came down to 10,909 U/L on the fourth day of admission. The boy had good urine output and there was no more passage of dark-colored urine since day three of the admission. Nasopharyngeal swab taken during admission was positive for respiratory syncytial virus. Dyspnea gradually resolved. On day five of illness, the CK, ALT and LDH had markedly improved. Repeated urine myoglobin and hemoglobin testing were negative. He was followed up two weeks later and was well then. There was no more passage of dark-colored urine.\n\nThereafter, he defaulted follow up and presented to us again at six years of age. This time, he complained of bilateral calf pain with antalgic gait associated with coryzal symptoms and fever. The episode was not preceded by any injuries nor excessive exercises. There was also no preceding administration of any intramuscular injection. Upon detailed history taking, the patient revealed that he frequently got muscle cramps after exercise. There was no family history of recurrent rhabdomyolysis, malignant hyperthermia or musculoskeletal diseases. The parents were both Chinese and non-consanguineous.\n\nPhysical examination revealed bilateral calf tenderness in the absence of weakness. There was no associated hepatosplenomegaly. He was well built with a weight and height at the 50th and 75th percentile, respectively, for his age and sex. Investigations during this admission again revealed an elevated CK (27,442 U/L), LDH (658 U/L) and ALT (138 U/L). Urine for myoglobin and hemoglobin (analyzed by same methods as before) were both positive. Plasma creatinine and potassium were normal. Hydration was instituted and the CK levels normalized 10 days later. Due to the recurrent episodes of rhabdomyolysis, further investigations were undertaken including blood gas, plasma lactate and ammonia but they were all unremarkable. Metabolic screening was performed and showed a normal pattern of plasma amino acids, serum acylcarnitine profile and urine organic acid profile. Genetic analysis of all coding exons of LPIN1 reviewed two heterozygous novel mutations: NM_145693.2(LPIN1):c.[1949_1967dupGTGTCACCACGCAGTACCA]; [2410G>C] (p.[Gly657Cysfs12];[Asp804His]) (Figure 1). The parents also underwent genetic tests for LPIN1. Each of the parents was a heterozygous carrier of one of the patient’s variants, which confirmed compound heterozygosity of the variants in the patient. The former variant was classified as likely pathogenic while the latter variant was classified as a variant of uncertain significance (VUS). Although the genetic findings were inconclusive, the patient’s clinical presentation was compatible with LPIN1-related acute recurrent rhabdomyolysis, and muscle biopsy was not done.\n\n(a) NM_145693.2(LPIN1):c.[1949_1967dupGTGTCACCACGCAGTACCA] p.[Gly657Cysfs12]. (b) NM_145693.2(LPIN1):c.[2410G>C] p.[Asp804His].\n\nIn spite of the uncertainty in the genetic findings, the likely diagnosis of LPIN-related acute recurrent rhabdomyolysis was conveyed to the family with proper genetic counseling conducted. Advice to avoid recurrent episodes of rhabdomyolysis, which include limiting strenuous exercise, ample hydration during exercise and prompt medical consultation in case of myalgia, was given. One year after the genetic diagnosis, he has been growing well and did not complain of passing any dark-colored urine, even during febrile illness.\n\nThe parents decided to have another baby and opted not for prenatal testing for LPIN1 variants. Prenatal testing carries intrinsic risk to the fetus and, in our opinion, is not recommended for this disease which is amendable to treatment and not debilitating. Post-natal genetic test on the younger sister confirmed that she also carried the two familial heterozygous variants; upon this finding, counselling was given to the parents.\n\n\nDiscussion\n\nLPIN1 mutations resulting in recurrent rhabdomyolysis was first described in three patients by Zeharia et al. in 200812. In less than a decade since it was first described, there has been an increasing evidence in the literature that LPIN1 mutations are one of the most frequent causes of recurrent rhabdomyolysis in childhood. Michot et al. reported that after excluding primary fatty acid oxidation disorders, recessive LPIN1 gene mutations were found in 59% of the patients who exhibited severe recurrent rhabdomyolysis in childhood. The male to female ratio was reported to be 0.89. Out of the 17 cases described by Michot et al., more than half of them were Caucasian and the others were of African, Asian (Vietnamese) and Maghrebi ethnicity13. Due to its relative novelty, the prevalence of the disease is still unknown. To our knowledge, there has been no previous report in the English-language literature of a child with Chinese ethnicity and LPIN1-related acute recurrent rhabdomyolysis (MIM #268200).\n\nThe pathophysiology of LPIN1-related recurrent rhabdomyolysis is not fully understood. Accumulation of lipid droplets is a common finding in muscle biopsy14. The loss of PAP enzymatic activity, rather than a loss in transcription co-regulator activity, plays a major role in the pathogenesis of this disease as the loss of the PAP activity alone without a loss in transcription co-regulator activity was sufficient to cause rhabodomyolysis15. Apart from the PAP activity responsible for triacylglycerol and phospholipid biosynthesis, it also serves as a transcriptional co-activator together with peroxisome proliferator activated receptor coactivator-1α (PPARA) and peroxisome proliferator activated receptor gamma coactivator-1 alpha (PGC-1α) to regulate genes encoding fatty acid oxidation and respiratory chain enzymes6,7. Multiple genes are either up-regulated or down-regulated in myoblasts of lipin-1-deficient patients16. Furthermore, various inflammatory inducers including lipopolysaccharides, zymosan and proinflammatory cytokines have been shown to repress LPIN1 expression in adipose tissue and muscular tissue17,18. Inflammatory inducers are usually produced when the body is under catabolic stress and the resulting suppressed lipin 1 activity may lead to rhabdomyolysis. This hypothesis might explain the fact that the majority of the severe episodes are associated with catabolic states such as a preceding illness13. Lipin 1 was also found to play in a role in autophagy of mitochondria as PAP activity is required for the maturation of autolysosomes and lipin 1 deficiency leads to accumulation of aberrant mitochondria19. This may explain the finding of mitochondrial aggregates on ultrastructural examination of muscle tissue in some patients14. Sarcoplasmic reticulum stress response was recently identified to be an important player in the pathogenesis of this disease by contributing to catastrophic de novo lipogenesis of both phospholipids and neutral lipids20.\n\nDisease-causing LPIN1 mutations are scattered throughout the coding region14. An intragenic deletion (c.2295-863_2410-27del) is frequently seen in Caucasian patients, possibly representing a founder effect13,14. In our patient, two heterozygous novel variants of LPIN1 gene are detected: the 19-nucleotide duplication NM_145693.2(LPIN1):c.[1949_1967dupGTGTCACCACGCAGTACCA] (p.[Gly657Cysfs12]) causes a frameshift with a premature termination codon and hence has deleterious effect; the novel missense variant, NM_145693.2(LPIN1):c.[2410G>C] (p.[Asp804His]), is found in the highly conserved LNS2 (Lipin/Ned1/Smp2) domain near the C terminus of lipin 1, which contains an important Mg2+-dependent catalytic site responsible for the PAP activity21. This amino acid residue is highly conserved. The variant is predicted to be deleterious (SIFT), damaging (PROVEAN) and probably damaging (Polyphen-2) by in silico analyses. At the time of reporting, both variants are absent from controls in the Exome Sequencing Project, 1000 Genomes Project, Exome Aggregation Consortium and Genome Aggregation databases. NM_145693.2(LPIN1):c.[1949_1967dupGTGTCACCACGCAGTACCA] (p.[Gly657Cysfs12]) was classified as a likely pathogenic variant (Criteria: PVS1 and PM2) and NM_145693.2(LPIN1):c.[2410G>C] (p.[Asp804His]) was classified as a variant of uncertain significance (VUS)(Criteria: PM2, PM3 and PP3) based on the American College of Medical Genetics and Genomics (ACMG) guideline published in 201522. Functional characterization of the two detected variants is warranted to confirm their pathogenicity.\n\nAll patients with LPIN1 mutations present with recurrent rhabdomyolysis triggered by febrile illness, prolonged fasting or anesthesia12,13. Michot et al. reviewed 29 cases with recurrent rhabdomyolysis, in which 17 patients were confirmed to harbor LPIN1 mutations. For the patients who harbor LPIN1 mutations, the first episodes of rhabdomyolysis occurred before the age of five years. The mean age of presentation was 21 months old and the earliest age of presentation was at five months of age. The number of recurrent episodes ranged from one to ten per patient13. Our patient presented at 15 months of age and has so far experienced two episodes of rhabdomyolysis. As more than 50% of patients with recurrent rhabdomyolysis were found to have LPIN1 mutations after exclusion of FAOD13, genetic testing for LPIN1 mutations should be performed prior to muscle biopsy if metabolic investigations are unremarkable. Muscle biopsy is an invasive procedure and require general anesthesia which may trigger rhabdomyolysis.\n\nMichot et al. also noted that among the 29 cases of unexplained rhabdomyolysis they studied, all cases eventually confirmed to have LPIN1 mutations had a CK level of >10,000 U/L13. In our patient, the CK levels were 127,494 U/L and 27,442 U/L during the first and second episodes of rhabdomyolysis, respectively. This is in line with Michot’s findings. Therefore, one should be vigilant to look for LPIN1 mutations in any young children presenting with rhabdomyolysis with a CK level of >10,000 U/L.\n\nThe management of rhabdomyolysis is mostly supportive by providing aggressive fluid replacement therapy to protect the renal function. Besides, close monitoring is also indispensable during fasting and anesthesia. As rhabdomyolysis episodes could be fatal, prevention of such events is of utmost importance. Parents should be well educated on the symptoms and precipitating factors of rhabdomyolysis. They should also be given an action plan when the child exhibits muscle pain or fever. In our case, proper education of the parents at diagnosis has helped in preventing further episodes of rhabdomyolysis. In a group of five patients with biallelic LPIN mutations, Pichler et al. proposed that high-caloric intake at periods of stress and intravenous glucose during episodes of rhabdomyolysis may decrease the number of rhabdomyolysis episodes and the duration of each episode by preventing catabolism22. Further studies are required to confirm the efficacy and safety of this approach.\n\nFor LPIN1-related acute recurrent rhabdomyolysis, it was reported that mortality was as high as 30% during an acute episode of severe rhabdomyolysis14. Between episodes, these patients thrived well. Intellectual disability was reported in one case at the age of nine years12. Our patient has been normal between the episodes apart from frequent muscle cramps after exercise and there is no evidence of intellectual disability. However, further observation is needed in view of the young age of our patient. The long-term prognosis of this condition has been reported in a 25-year-old female with LPIN1-related recurrent rhabdomyolysis, who had bilateral common peroneal neuropathies in addition to a background residual distal myopathy detected one year following discharge from intensive care. It was uncertain whether the neuropathies were a result of critical illness, compression and/or severe weight loss during the admission or intrinsic to the underlying genetic lipin 1 deficiency23.\n\nGenetic confirmation of LPIN1 mutations not only helps the patient but also their family members. In our case, the younger sister had genetic testing done soon after birth. This has raised the parents’ awareness of the disease and its preventive measures. The younger sister has been growing well and has had no episodes of rhabdomyolysis at the time of reporting at one year of age.\n\n\nConclusion\n\nThis is the first probable reported case of LPIN1-related rhabdomyolysis in the Hong Kong Chinese population. It has been shown that LPIN1-related rhabdomyolysis is a major cause of severe rhabdomyolysis in early childhood so we should be vigilant of this condition. Screening for LPIN1 mutations should be undertaken at an early stage in the work up of childhood rhabdomyolysis. As rhabdomyolysis in children has a high mortality rate, early recognition of the symptoms, timely diagnosis and proactive management are essential. Regarding prognosis, more observations and longitudinal studies are warranted to determine the long-term outcomes of this condition.\n\n\nData availability\n\nNCBI ClinVar: NM_145693.2(LPIN1):c.[1949_1967dupGTGTCACCACGCAGTACCA] (p.[Gly657Cysfs12]). Accession number SCV000965694.\n\nNCBI ClinVar: NM_145693.2(LPIN1):c.[2410G>C] (p.[Asp804His]). Accession number SCV000965695.\n\n\nConsent\n\nWritten informed consent for publication of the clinical details was obtained from the parents of the patient.", "appendix": "References\n\nPéterfy M, Phan J, Xu P, et al.: Lipodystrophy in the fld mouse results from mutation of a new gene encoding a nuclear protein, lipin. Nat Genet. 2001; 27(1): 121–124. PubMed Abstract \| Publisher Full Text\n\nDonkor J, Sariahmetoglu M, Dewald J, et al.: Three mammalian lipins act as phosphatidate phosphatases with distinct tissue expression patterns. J Biol Chem. 2007; 282(6): 3450–3457. PubMed Abstract \| Publisher Full Text\n\nDonkor J, Zhang P, Wong S, et al.: A conserved serine residue is required for the phosphatidate phosphatase activity but not the transcriptional coactivator functions of lipin-1 and lipin-2. J Biol Chem. 2009; 284(43): 29968–29978. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMelé M, Ferreira PG, Reverter F, et al.: Human genomics. The human transcriptome across tissues and individuals. Science. 2015; 348(6235): 660–665. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHan GS, Wu WI, Carman GM: The Saccharomyces cerevisiae Lipin homolog is a Mg2+-dependent phosphatidate phosphatase enzyme. J Biol Chem. 2006; 281(14): 9210–9218. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFinck BN, Gropler MC, Chen Z, et al.: Lipin 1 is an inducible amplifier of the hepatic PGC-1alpha/PPARalpha regulatory pathway. Cell Metab. 2006; 4(3): 199–210. PubMed Abstract \| Publisher Full Text\n\nZhang P, Takeuchi K, Csaki LS, et al.: Lipin-1 phosphatidic phosphatase activity modulates phosphatidate levels to promote peroxisome proliferator-activated receptor γ (PPARγ) gene expression during adipogenesis. J Biol Chem. 2012; 287(5): 3485–3494. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhang R, Jiang F, Hu C, et al.: Genetic variants of LPIN1 indicate an association with Type 2 diabetes mellitus in a Chinese population. Diabet Med. 2013; 30(1): 118–122. PubMed Abstract \| Publisher Full Text\n\nLoos RJ, Rankinen T, Pérusse L, et al.: Association of lipin 1 gene polymorphisms with measures of energy and glucose metabolism. Obesity (Silver Spring). 2007; 15(11): 2723–2732. PubMed Abstract \| Publisher Full Text\n\nWiedmann S, Fischer M, Koehler M, et al.: Genetic variants within the LPIN1 gene, encoding lipin, are influencing phenotypes of the metabolic syndrome in humans. Diabetes. 2008; 57(1): 209–217. PubMed Abstract \| Publisher Full Text\n\nFawcett KA, Grimsey N, Loos RJ, et al.: Evaluating the role of LPIN1 variation in insulin resistance, body weight, and human lipodystrophy in U.K. Populations. Diabetes. 2008; 57(9): 2527–2533. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZeharia A, Shaag A, Houtkooper RH, et al.: Mutations in LPIN1 cause recurrent acute myoglobinuria in childhood. Am J Hum Genet. 2008; 83(4): 489–494. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMichot C, Hubert L, Brivet M, et al.: LPIN1 gene mutations: a major cause of severe rhabdomyolysis in early childhood. Hum Mutat. 2010; 31(7): E1564-1573. PubMed Abstract \| Publisher Full Text\n\nMichot C, Hubert L, Romero NB, et al.: Study of LPIN1, LPIN2 and LPIN3 in rhabdomyolysis and exercise-induced myalgia. J Inherit Metab Dis. 2012; 35(6): 1119–1128. PubMed Abstract \| Publisher Full Text\n\nSchweitzer GG, Collier SL, Chen Z, et al.: Rhabdomyolysis-Associated Mutations in Human LPIN1 Lead to Loss of Phosphatidic Acid Phosphohydrolase Activity. JIMD Rep. 2015; 23: 113–122. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMichot C, Mamoune A, Vamecq J, et al.: Combination of lipid metabolism alterations and their sensitivity to inflammatory cytokines in human lipin-1-deficient myoblasts. Biochim Biophys Acta. 2013; 1832(12): 2103–2114. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFeingold KR, Moser A, Patzek SM, et al.: Infection decreases fatty acid oxidation and nuclear hormone receptors in the diaphragm. J Lipid Res. 2009; 50(10): 2055–2063. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLu B, Lu Y, Moser AH, et al.: LPS and proinflammatory cytokines decrease lipin-1 in mouse adipose tissue and 3T3-L1 adipocytes. Am J Physiol Endocrinol Metab. 2008; 295(6): E1502–E1509. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhang P, Verity MA, Reue K: Lipin-1 regulates autophagy clearance and intersects with statin drug effects in skeletal muscle. Cell Metab. 2014; 20(2): 267–279. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRashid T, Nemazanyy I, Paolini C, et al.: Lipin1 deficiency causes sarcoplasmic reticulum stress and chaperone-responsive myopathy. EMBO J. 2019; 38(1): pii: e99576. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBurroughs AM, Allen KN, Dunaway-Mariano D, et al.: Evolutionary genomics of the HAD superfamily: understanding the structural adaptations and catalytic diversity in a superfamily of phosphoesterases and allied enzymes. J Mol Biol. 2006; 361(5): 1003–1034. PubMed Abstract \| Publisher Full Text\n\nPichler K, Scholl-Buergi S, Birnbacher R, et al.: A novel therapeutic approach for LPIN1 mutation-associated rhabdomyolysis--The Austrian experience. Muscle Nerve. 2015; 52(3): 437–439. PubMed Abstract \| Publisher Full Text\n\nStepien KM, Schmidt WM, Bittner RE, et al.: Long-term outcomes in a 25-year-old female affected with lipin-1 deficiency. JIMD Rep. 2019; 46(1): 4–10. PubMed Abstract \| Free Full Text" }	[ { "id": "53286", "date": "16 Sep 2019", "name": "Michael F. Buckley", "expertise": [ "Reviewer Expertise Clinical genomics." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is the initial description of compound heterozygous variants in LPIN1 as a cause of rhadomyolysis in a person of Chinese ethnicity.\nThe sequencing data provided is robust, the nomenclature used is correct and the ACMG classification for the missense change is conservative.\n\nThere is little in the way of improvements to the manuscript that I could suggest. The categorical statement \"The human LPIN1 gene is mapped to chromosome 2p25.1 and contains 20 exons\" isn't correct as this gene has multiple transcription isoforms with different numbers and combinations of exons. The isoform used in this discussion certainly has 20 exons but other isoforms range up to 22 exons.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [] }, { "id": "54678", "date": "31 Oct 2019", "name": "Brian N. Finck", "expertise": [ "Reviewer Expertise Intermediary metabolism" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nLipin 1 plays important roles in regulating intermediary metabolism by acting as a lipid phosphatase to synthesize diacylglycerol from phosphatidic acid. This is an important step in glycerolipid synthesis and a metabolic branch point. Mutations in lipin 1 in people lead to severe, recurrent rhabdomyolysis that can be debilitating or fatal. While these mutations are rare in the general population, they account for a significant proportion of acute paediatric rhabdomyolysis cases. In this case report, Yim and colleagues detail the identification of novel mutations in LPIN1 that are associated with rhabdomyolysis in a boy in Hong Kong. The report is well done and complete. The authors have done a good job of detailing the important literature on the topic. No changes are requested.\n\nIs the background of the case’s history and progression described in sufficient detail? No\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1566
https://f1000research.com/articles/8-1565/v1	02 Sep 19	{ "type": "Research Article", "title": "The frequency of epithelial ovarian cancer subtypes in Sudanese women at Omdurman Maternity Hospital, 2013-2018: A cross-sectional study", "authors": [ "Rawia Eljaili Elmassry", "Nassr Eldin M.A. Shrif", "Aisha Osman Mohamed", "Fayad Jamaleldin", "Arwa Elaagip", "Nazik Elmalaika Husain", "Nassr Eldin M.A. Shrif", "Aisha Osman Mohamed", "Fayad Jamaleldin", "Arwa Elaagip", "Nazik Elmalaika Husain" ], "abstract": "Background: Globally, epithelial ovarian carcinoma (EOC) is considered the gynecological cancer with the highest mortality. In Sudan, there are scarce publications about the frequency of this carcinoma. Therefore, the present study intended to perform a cross-sectional study to review the morphological sub-types and sort EOC according to age and grade in Omdurman Maternity Hospital (OMH) in Sudan. Methods: This cross-sectional, hospital-based study included 70 EOC cases diagnosed at OMH in the period 2013-2018. The data were collected from OMH records in the period 2016-2018, and included ovarian cancer types, ages of patients, and tumor grades. Results: The participants’ median age was 50 years, and the majority of EOC cases were in younger patients (48.6%; n=34; ≤ 50 years (18 to 50 years)). The most familiar tumor sub-type was serous carcinoma (44.3%; n=31), followed by endometrioid carcinoma (27.1%; n=19), mucinous carcinoma (17.1%; n=12), clear cell carcinoma (8.6%; n=6) and undifferentiated carcinoma (2.9%; n=2). The majority of cases were categorized as low grade (51.4%; n=36). Our results revealed significant relationships between EOC types and grades (Fisher’s Exact test, p=0.000). Conclusion: In Sudanese patients with EOC, serous carcinoma is the most common histological subtype, and EOC is likely to occur in women of a younger age (<50 years). Our results indicate a younger presentation of EOC and warrants quick and thorough investigation of any vague abdominal complaint in women of a younger age (<50 years). Also, it may help in guiding researchers developing screening programs especially for younger women, pay attention to the serous type as the common type and finding novel biomarkers especially for treatment and prognosis of this type.", "keywords": [ "Epithelial Ovarian Cancer", "Sudan", "hospital-based", "Omdurman Maternity Hospital" ], "content": "Introduction\n\nOvarian carcinoma (OC) is the second most frequent gynecological cancer in developing countries and globally1,2. They are the fifth most mortal disease in women3,4, and in 2018, ~22,240 new events and 14,070 deaths were associated with OCs in the United States5,6. In Sudan, OCs ranked as the fourth highest malignancy in women, with an incidence rate of ~188 for every 100,000 people, which is 8 for every 100,000 people when gender is specified, and 7 for every 100,000 people with an age-standardized rate7,8.\n\nEpithelial ovarian carcinoma (EOC) is the most widespread and has the highest mortality of OC1,9, with high-grade serous ovarian cancer classed as the deadliest sub-type10,11. In the United State, generally, EOC has been considered an age-related disease; women aged >65 years are considered the target population12,13. Although the incidence of OCs is decreasing, in the majority of cases, distant metastasis is seen at diagnosis14,15. Surgery and chemotherapy are the most important treatments of this cancer but are efficient only in the early stages16,17. Currently; the prognosis for this cancer is still poor, and the five-year survival rate remains low18,19. In Sudan, scarce publications are available about the frequency and incidence of EOC. The present study intended to evaluate the morphological sub-types in Sudan using data from one hospital, sort the disease by age and grade and compare the results with similar studies in other regions of Sudan and other countries.\n\n\nMethods\n\nThis was a cross-sectional, hospital-based study performed to determine the frequency of newly diagnosed EOC subtypes, as well as to identify the relationship between subtypes and tumor grades and ages of patients.\n\nThe Ethics Committee of Alzaiem Alazhari University and Omdurman Maternity Hospital approved the study and waived informed consent from patients since there was anonymization of patients’ identity and only laboratory numbers were used.\n\nA total of 70 EOC patients from Omdurman Maternity Hospital diagnosed in the period between 2013 and 2018 were selected for this study via convenience sampling. The data were collected in the period between 2016 and 2018. All patients were adult women with OC and no other malignancies. Patients with tiny biopsies, missed blocks, or missing data (more than 50%) were excluded.\n\nData were collected from hospital records, which included patients’ ages, EOC types, and tumor grades.\n\nThe statistical analysis of the collected data was performed using SPSS, version 24 (IBM SPSS). Descriptive analysis was performed for all study variables: EOC types, patient’s ages, and tumor grades. Fisher’s exact test was used for finding statistical significance between EOC types, patients’ ages, and tumor grades. A p-value of <0.05 was considered to be statistically significant.\n\n\nResults\n\nThis study was conducted among 70 new EOC cases. The median age of participants was 50 years (range, 18 – 75 years). In total, 48.6% (n=34) of the participants were in the age group ≤ 50 years(18 to 50 years), and 44.3% (n=31) were in the age group > 50 years (51 to 75 years).The frequency of common histological subtypes is shown in Figure 1. The results showed that the majority of the cases were low grade (51.4%; n=36) compared with high grade (48.6%; n=34).\n\nOur results showed significant associations between EOC tumor types and EOC tumor grades (Fisher’s Exact test, p = 0.000) and no significant associations between tumor types and ages (Fisher’s Exact test, p>0.05).\n\n\nDiscussion\n\nOCs are the most lethal gynecological malignancy worldwide and constitutes the fifth leading cause of cancer-related death in women20,21. Comparison of the age distribution of new OC cases revealed that the median age of 50 years recorded in this study is similar to a local study from Gezira State, Sudan7, sub-Saharan African22,23 and other developing countries24. Nevertheless, it is lower than that reported from developed countries25, for example a median age of 65 years was reported from the Netherlands26. The lower age in this study may suggest an earlier onset or a more aggressive course of the disease in our environment. Also, it may be because Africa has the youngest population of any of the continents27. Our study results indicate a younger presentation of EOC and warrants quick and thorough investigation of any vague abdominal complaint in women of a younger age (<50 years). In the present study, serous adenocarcinoma is the most frequent histological type (43.1%), which is supported by previous studies28–30.\n\nThis research was affected by the inherent limitations of the retrospective type of study. Also, it was reliant on the data extracted from medical records, was based at a single institution-based, and had a small sample size. Finally, there were missing variables for some participants; incomplete or missing documents may lead to the introduction of bias into the study and affect the outcome.\n\n\nConclusion\n\nThe present study revealed that contrary to some previously published data, EOC affected a larger proportion of younger Sudanese women. The most prevalent subtype of EOC is serous carcinoma. A further study of frequency and prevalence of EOCs in other parts of Sudan is needed.\n\n\nData availability\n\nFigshare: frequency data.csv, https://doi.org/10.6084/m9.figshare.925520031\n\nData are available under the terms of the Creative Commons Zero “No rights reserved” data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nThe authors confirm that no grants were involved in supporting this work.\n\n\nAcknowledgments\n\nWe thank the Department of Histopathology and Cytology at Omdurman Maternity Hospital for providing their facilities. We are also thankful to our technician Mr. Hassan for his cooperation during data collection.\n\n\nReferences\n\nLu DH, Yang J, Gao LK, et al.: Lysine demethylase 2A promotes the progression of ovarian cancer by regulating the PI3K pathway and reversing epithelial‑mesenchymal transition. Oncol Rep. 2019; 41(2): 917–927. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEl-Kott AF, Shati AA, Al-Kahtani MA, et al.: Acylated Ghrelin Renders Chemosensitive Ovarian Cancer Cells Resistant to Cisplatin Chemotherapy via Activation of the PI3K/Akt/mTOR Survival Pathway. Anal Cell Pathol (Amst). 2019; 2019: 9627810. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKouba S, Ouldamer L, Garcia C, et al.: Lipid metabolism and Calcium signaling in epithelial ovarian cancer. Cell Calcium. 2019; 81: 38–50. PubMed Abstract \| Publisher Full Text\n\nCeccarelli S, Megiorni F, Bellavia D, et al.: Notch3 Targeting: A Novel Weapon against Ovarian Cancer Stem Cells. Stem Cells Int. 2019; 2019: 6264931. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJiang S, Chang H, Deng S, et al.: Icariin enhances the chemosensitivity of cisplatin‑resistant ovarian cancer cells by suppressing autophagy via activation of the AKT/mTOR/ATG5 pathway. Int J Oncol. 2019; 54(6): 1933–1942. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhao Y, Cao J, Melamed A, et al.: Losartan treatment enhances chemotherapy efficacy and reduces ascites in ovarian cancer models by normalizing the tumor stroma. Proc Natl Acad Sci. 2019; 116(6): 2210–2219. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAbuidris DO, Weng HY, Elhaj AM, et al.: Incidence and survival rates of ovarian cancer in low-income women in Sudan. Mol Clin Oncol. 2016; 5(6): 823–828. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAdam W, Gurashi RA, Humida MA, et al.: Ovarian Cancer in Sudan. J Med Biol Sci Res. 2017; 3(4): 37–41. Reference Source\n\nGong Y, Fan X: MiR-539-3p promotes the progression of epithelial ovarian cancer by targeting SPARCL1. Eur Rev Med Pharmacol Sci. 2019; 23(6): 2366–2373. PubMed Abstract \| Publisher Full Text\n\nColombo N, Sessa C, du Bois A, et al.: ESMO-ESGO consensus conference recommendations on ovarian cancer: pathology and molecular biology, early and advanced stages, borderline tumours and recurrent disease†. Ann Oncol. 2019; 30(5): 672–705. PubMed Abstract \| Publisher Full Text\n\nSavant SS, Sriramkumar S, O’Hagan H: The Role of Inflammation and Inflammatory Mediators in the Development, Progression, Metastasis, and Chemoresistance of Epithelial Ovarian Cancer. Cancers (Basel). 2018; 10(8): pii: E251. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChon HS, Sehovic M, Marchion D, et al.: Biologic Mechanisms Linked to Prognosis in Ovarian Cancer that May Be Affected by Aging. J Cancer. 2019; 10(12): 2604. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDeng F, Xu X, Lv M, et al.: Age is associated with prognosis in serous ovarian carcinoma. J Ovarian Res. 2017; 10(1): 36. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKroeger PT Jr, Drapkin R: Pathogenesis and heterogeneity of ovarian cancer. Curr Opin Obstet Gynecol. 2017; 29(1): 26–34. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiu Y, Lin J, Zhai S, et al.: MicroRNA-214 Suppresses Ovarian Cancer by Targeting β-Catenin. Cell Physiol Biochem. 2018; 45(4): 1654–1662. PubMed Abstract \| Publisher Full Text\n\nGasparri ML, Bardhi E, Ruscito I, et al.: PI3K/AKT/mTOR Pathway in Ovarian Cancer Treatment: Are We on the Right Track? Geburtshilfe Frauenheilkd. 2017; 77(10): 1095–1103. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nde Melo AC, Paulino E, Garces ÁH: A Review of mTOR Pathway Inhibitors in Gynecologic Cancer. Oxid Med Cell Longev. 2017; 2017: 4809751. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChong T, Sarac A, Yao CQ, et al.: Deregulation of the spindle assembly checkpoint is associated with paclitaxel resistance in ovarian cancer. J Ovarian Res. 2018; 11(1): 27. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nReid BM, Permuth JB, Sellers TA: Epidemiology of ovarian cancer: a review. Cancer Biol Med. 2017; 14(1): 9–32. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiu H, Liu J, Wang S, et al.: Enterolactone has stronger effects than enterodiol on ovarian cancer. J Ovarian Res. 2017; 10(1): 49. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSakhuja S, Yun H, Pisu M, et al.: Availability of healthcare resources and epithelial ovarian cancer stage of diagnosis and mortality among Blacks and Whites. J Ovarian Res. 2017; 10(1): 57. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVanderpuye V, Yarney J: Ovarian cancer: an analysis of forty-four patients at the National Radiotherapy Centre, Accra--Ghana. West Afr J Med. 2007; 26(2): 93–96. PubMed Abstract\n\nRabiu KA, Akinola OI, Adewunmi AA, et al.: Delays in presentation and management of ovarian cancer in Lagos, Nigeria. J Obstet Gynaecol. 2013; 33(3): 305–308. PubMed Abstract \| Publisher Full Text\n\nBasu P, De P, Mandal S, et al.: Study of 'patterns of care' of ovarian cancer patients in a specialized cancer institute in Kolkata, eastern India. Indian J Cancer. 2009; 46(1): 28–33. PubMed Abstract \| Publisher Full Text\n\nQuirk JT, Natarajan N: Ovarian cancer incidence in the United States, 1992-1999. Gynecol Oncol. 2005; 97(2): 519–523. PubMed Abstract \| Publisher Full Text\n\nvan Altena AM, Karim-Kos HE, de Vries E, et al.: Trends in therapy and survival of advanced stage epithelial ovarian cancer patients in the Netherlands. Gynecol Oncol. 2012; 125(3): 649–654. PubMed Abstract \| Publisher Full Text\n\nHarford JB: Breast-cancer early detection in low-income and middle-income countries: do what you can versus one size fits all. Lancet Oncol. 2011; 12(3): 306–312. PubMed Abstract \| Publisher Full Text\n\nKrishnaswamy P, Nayak A, Shivananjiah C, et al.: Study of clinical presentation and histopathological patterns of ovarian cancer in a tertiary care centre. Int J Community Med Public Health. 2017; 3(1): 86–89. Publisher Full Text\n\nPradhan HK, Singh P, Ravikumar MS, et al.: Study of risk factors and tumor markers in ovarian malignancy in western part of Odisha: a prospective observational study. Int J Reprod Contracept Obstet Gynecol. 2018; 7(4): 1571–1578. Publisher Full Text\n\nCho KR, Shih IeM: Ovarian cancer. Annu Rev Pathol. 2009; 4: 287–313. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nelmassry r, Shrif NEMA, Mohammed AO, et al.: frequancy data.csv. figshare. Dataset. 2019. http://www.doi.org/10.6084/m9.figshare.9255200.v2" }	[ { "id": "53798", "date": "01 Oct 2019", "name": "Ahmed Abdalla Mohamedani", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a concise piece of work and perhaps the third paper addressing ovarian cancer in Sudan. The title is concise and reflects very well the content of the research, but a minor change below may be adopted by the authors:\n“The frequency of epithelial ovarian cancer subtypes in Sudanese women at Omdurman Maternity Hospital in the period 2013-2018, Sudan”\nComments\nMethods:\nThe authors may agree to add the word retrospective to the type of the study.\n\nHave the authors designed a data collection form for the purpose.\n\nHave they revised some of the archived histology slides or produced some from the archived wax blocks.\n\nWhat is the total number of EOC diagnosed during the period 2013- 2018 and what is thus the number excluded according to the exclusion criteria set by the authors.\n\nWhat is meant by “newly diagnosed EOC subtypes “diagnosed or described as this is a retrospective study and there are no new cases. (line 2 under study design)\n\nResults:\nOnly one figure is presented. What about other data regarding residence, parity etc.\n\nIt may be a good idea if the authors would agree to add a table/figure to display the tumour grades.\nDiscussion:\nIs very short compared to the number of cited references.\nReferences:\n\nAre adequate and covered the research subject quite well but:\nAbout a third of the references covered the pathogenesis, behaviour, response to therapy and molecular biology of EOC.\n\nReference 27 (please see below) looks a bit odd as it is about breast cancer. I know why it is cited (young African population) but still it looks odd.\n27. Harford JB: Breast-cancer early detection in low-income and middle-income countries: do what you can versus one size fits all. Lancet Oncol. 2011; 12(3): 306–312. PubMed Abstract \| Publisher Full Text\n\nI have also annotated a copy of the manuscript with some comments and corrections to the text.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "57732", "date": "09 Jan 2020", "name": "Sulma Ibrahim Mohammed", "expertise": [ "Reviewer Expertise Cancer research and interest in cancer in Sudan." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article describes the frequency of epithelial ovarian cancer subtypes in Omdurman Maternity Hospital. It examines the association of EOC with age and determined subtypes and grade association. The study is significant as more cancer documentation is needed in Sudan. That said my enthusiasm for this study greatly hampered by the following:\nOmdurman Maternity Hospital is not the major hospital where women with cancers are diagnosed - so that may affect the conclusion.\n\nNot sure from where the authors extracted the tissue grading - as this is not a common practice in Sudan and therefore the authors need to provide more details and present results with some pictures.\n\nDetail is also needed in how the data for subtype is obtained - from pathology reports or by using archival blocks and then repeat the pathology reports?\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-1565
https://f1000research.com/articles/8-188/v1	15 Feb 19	{ "type": "Data Note", "title": "A curated transcriptome dataset collection to investigate inborn errors of immunity", "authors": [ "Salim Bougarn", "Sabri Boughorbel", "Damien Chaussabel", "Nico Marr", "Sabri Boughorbel", "Damien Chaussabel", "Nico Marr" ], "abstract": "Primary immunodeficiencies (PIDs) are a heterogeneous group of inherited disorders, frequently caused by loss-of-function and less commonly by gain-of-function mutations, which can result in susceptibility to a broad or a very narrow range of infections but also in inflammatory, allergic or malignant diseases. Owing to the wide range in clinical manifestations and variability in penetrance and expressivity, there is an urgent need to better understand the underlying molecular, cellular and immunological phenotypes in PID patients in order to improve clinical diagnosis and management. Here we have compiled a manually curated collection of public transcriptome datasets mainly obtained from human whole blood, peripheral blood mononuclear cells (PBMCs) or fibroblasts of patients with PIDs and of control subjects for subsequent meta-analysis, query and interpretation. A total of nineteen (19) datasets derived from studies of PID patients were identified and retrieved from the NCBI Gene Expression Omnibus (GEO) database and loaded in GXB, a custom web application designed for interactive query and visualization of integrated large-scale data. The dataset collection includes samples from well characterized PID patients that were stimulated ex vivo under a variety of conditions to assess the molecular consequences of the underlying, naturally occurring gene defects on a genome-wide scale. Multiple sample groupings and rank lists were generated to facilitate comparisons of the transcriptional responses between different PID patients and control subjects. The GXB tool enables browsing of a single transcript across studies, thereby providing new perspectives on the role of a given molecule across biological systems and PID patients. This dataset collection is available at http://pid.gxbsidra.org/dm3/geneBrowser/list.", "keywords": [ "Transcriptomics", "microarray", "primary immunodeficiency disorders", "inborn errors of immunity." ], "content": "Introduction\n\nPrimary immunodeficiencies (PIDs) are a heterogeneous group of inherited disorders, most often caused by loss-of-function mutations and less commonly by gain-of-function mutations, affecting components of the innate and/or adaptive immune system1–3. These inborn errors of immunity can result in profoundly increased susceptibility to a broad or a very narrow range of infections but also autoimmune disorders, allergies and malignancies1,4–6. The spectrum of clinical manifestations of PIDs is very broad and largely dependent upon the affected gene(s) and the degree to which normal gene function is lost or altered. In addition, a variety of other factors such as germline or somatic mosaicism, modifier genes and environmental factors can play an important role in the clinical penetrance and expressivity of a given disease phenotype3,7,8. To date, mutations in more than 300 genes have been identified to cause PIDs, which are classified into major groups reflecting the diverse immunological phenotypes5,9. Nonetheless, PIDs often go unrecognized or are not properly diagnosed10. However, with the recent developments and rapidly declining costs of next-generation sequencing technologies and other high-throughput methods, it is expected that many more as yet unknown disease-related genetic variants will be discovered in the near future7.\n\nA considerable challenge for identifying causal genetic variants—which is critical for the diagnosis and clinical management of PID patients—lies in the vast heterogeneity of the underlying immunological phenotypes and clinical manifestations on the one hand and in the degree of human genetic variation between individuals on the other hand. Despite considerable advances in recent years, specific gene functions in humans, their roles and regulation in biological processes, and essentiality of the redundancy in a function of a particular gene for protective immunity of the human host remains poorly understood6. In many cases, the use of forward and reverse genetics in mice or other model organisms has provided insufficient insights into the pathophysiology of PIDs, this due to interspecies differences and the fact that inbreeding has let to various deficiencies in laboratory animals, rendering them susceptible to a broad range of infections that often poorly recapitulates the clinical phenotypes in humans11. On the other hand, studying naturally occurring genetic defects in humans is much more challenging given the difficulty of obtaining biological samples, ethical implications and potential risks that go along with it. An additional challenge is the low frequency of most null alleles. Although PIDs are not necessarily rare when considered collectively, the small number of individuals that suffer from a specific deficiency usually does not permit classic case-control or family-based genetic association studies. Indeed, a considerable proportion of monogenic etiologies of PIDs were initially reported in single patients12. The ability to identify single-gene inborn errors in PID patients requires validation of the disease-causing variant by in-depth mechanistic studies demonstrating the structural and functional consequences of the mutations using blood or other accessible biological samples such as fibroblasts from skin biopsies12. In this context, several transcriptomics studies have been conducted using whole blood, PBMCs and fibroblasts of well-characterized PID patients, to assess the underlying immunological phenotypes at the molecular and cellular levels in more detail (Table 1) and in many cases, to further validate the causal relationship between the underlying genotypes and clinical phenotypes. Notable are several seminal studies of PID patients with susceptibility to a very narrow range of pathogens, such as patients with MYD88 or IRAK4 deficiency who are primarily susceptible to pyrogenic bacterial infections13,14, patients with TBK1, TRIF or TLR3 deficiency15–17 which underlies herpes simplex encephalitis of childhood, or a recent study of a child with IRF7 deficiency who was primarily susceptible to severe influenza but otherwise immunocompetent with regard to other common infectious diseases18. Such studies have highlighted that often, the underlying gene defect may only affect a narrow repertoire of transcriptional responses while the affected individual's cells remain highly responsive to specific stimulation through alternate receptors, pathways and signaling networks and in particular to ex vivo stimulation with whole organisms, reflecting the high degree of human gene redundancy in host defenses6.\n\nHere, we compiled a curated collection of 19 transcriptome datasets, retrieved from the NCBI's Gene Expression Omnibus (GEO) database, to provide as resource for the investigation on inborn errors of immunity. The datasets were loaded into a custom interactive web application, the Gene Expression Browser (GXB), (http://pid.gxbsidra.org/dm3/geneBrowser/list), which allows seamless access to the data and interactive visualization of the transcriptional responses, along with demographic and clinical information19. The user can customize data plots by adding multiple layers of parameters (e.g. age, gender, sample type and type of genetic defects), select and modify the sample ordering and gene rank lists, and generate links (mini URL) that can be shared via e-mail or used in publications. The GXB tool enables browsing of a single transcript across multiple studies and datasets, providing new perspectives on the role of a given molecule across biological systems and PID patients. In summary, this dataset collection can aid clinicians and researchers to study and quickly visualize the functional consequences of a variety of well characterized, naturally occurring mutations on a genome-wide scale.\n\n\nMethods\n\nA total of 157 datasets were identified in GEO using the following search query: Homo sapiens AND (“primary immunodeficiency diseases” OR “primary immunodeficiencies diseases” OR PID OR “autosomal recessive” OR “autosomal dominant” OR “inherited deficiency”) AND (\"Expression profiling by array”). All GEO entries that were returned with this query were manually curated. This process involved reading all the descriptions available for the datasets, the study designs and corresponding research articles. Finally, a total of 19 datasets were retained because they contained samples which were obtained from well characterized patients with known PIDs (i.e. the genetic etiology had been identified) or the samples were obtained from patients which were considered to have common variable immunodeficiency (CVID). These include datasets that were generated from whole blood, PBMCs, fibroblasts, B cells, iPS and kidney biopsy of individuals with defects in intrinsic and innate immunity, combined immunodeficiencies with associated or syndromic features, autoinflammatory disorders, congenital defects of phagocyte number, functions, or both, predominantly antibody deficiencies and diseases of immune dysregulation. The selected datasets are listed in Table 1. A breakdown of the dataset collection by category in accordance to the most recently published update on PID classification from International Union of Immunological Societies Expert Committee5,9 is shown in Figure 1.\n\nThe pie chart indicates the PIDs classification out for the 19 datasets.\n\nThe selected datasets were downloaded from GEO using the SOFT file format. Then, the datasets were uploaded onto our web tool, called the Gene Expression Browser (GXB), an interactive application hosted on the Amazon Web Services cloud19. Information about samples and study design were also uploaded. The available samples were assigned to groups based on the individuals and deficiencies studied and genes were ranked according to different group comparisons allowing the identification of transcripts that were differentially expressed between the patient's and control subject's cells cultured or stimulated ex vivo under the same conditions. Our dataset collection, uploaded in GXB, is available at http://pid.gxbsidra.org/dm3/geneBrowser/list. A web tutorial for the use of GXB can be accessed at: https://gxb.benaroyaresearch.org/dm3/tutorials.gsp#gxbtut.\n\nA detailed description of GXB has been recently published19,28–30 and is reproduced here so that readers can use this article as a standalone resource. Briefly, datasets of interest can be quickly identified either by filtering on criteria from pre-defined sections on the left or by entering a query term in the search box at the top of the dataset navigation page. Clicking on one of the studies listed in the dataset navigation page opens a viewer designed to provide interactive browsing and graphic representations of large-scale data in an interpretable format. This interface is designed to present ranked gene lists and display expression results graphically in a context-rich environment. Selecting a gene from the rank ordered list on the left of the data-viewing interface will display its expression values graphically in the screen’s central panel. Directly above the graphical display drop down menus give users the ability: a) To change how the gene list is ranked - this allows the user to change the method used to rank the genes, or to only include genes that are selected for specific biological interest; b) To change sample grouping (Group Set button) - in some datasets, a user can switch between groups based on cell type to groups based on disease type, for example; c) To sort individual samples within a group based on associated categorical or continuous variables (e.g. gender or age); d) To toggle between the bar chart view and a box plot view, with expression values represented as a single point for each sample. Samples are split into the same groups whether displayed as a bar chart or box plot; e) To provide a color legend for the sample groups; f) To select categorical information that is to be overlaid at the bottom of the graph - for example, the user can display gender or smoking status in this manner; g) To provide a color legend for the categorical information overlaid at the bottom of the graph; h) To download the graph as a portable network graphics (png) image. Measurements have no intrinsic utility in absence of contextual information. It is this contextual information that makes the results of a study or experiment interpretable. It is therefore important to capture, integrate and display information that will give users the ability to interpret data and gain new insights from it. We have organized this information under different tabs directly above the graphical display. The tabs can be hidden to make more room for displaying the data plots, or revealed by clicking on the blue “show info panel” button on the top right corner of the display. Information about the gene selected from the list on the left side of the display is available under the “Gene” tab. Information about the study is available under the “Study” tab. Rolling the mouse cursor over a bar chart feature while displaying the “Sample” tab lists any clinical, demographic, or laboratory information available for the selected sample. Finally, the “Downloads” tab allows advanced users to retrieve the original dataset for analysis outside this tool. It also provides all available sample annotation data for use alongside the expression data in third party analysis software. Other functionalities are provided under the “Tools” drop-down menu located in the top right corner of the user interface. Some of the notable functionalities available through this menu include: a) Annotations, which provides access to all the ancillary information about the study, samples and dataset organized across different tabs; b) Cross-project view; which provides the ability for a given gene to browse through all available studies; c) Copy link, which generates a mini-URL encapsulating information about the display settings in use and that can be saved and shared with others (clicking on the envelope icon on the toolbar inserts the URL in an email message via the local email client); d) Chart options; which gives user the option to customize chart labels.\n\n\nData availability\n\nAll datasets included in our curated collection are available publicly via the NCBI GEO website: https://www.ncbi.nlm.nih.gov/gds/ and are referenced throughout the manuscript by their GEO accession numbers (e.g. GSE92466). Signal files and sample description files can also be downloaded from the GXB tool under the “downloads” tab.", "appendix": "Grant information\n\nAll the authors listed on this publication received support from the Qatar Foundation. Support for this project was provided by the Qatar National Research Fund award NPRP10-0205-170348.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe would like to thank all the investigators who decided to make their datasets publicly available by depositing them in GEO.\n\n\nReferences\n\nSeleman M, Hoyos-Bachiloglu R, Geha RS, et al.: Uses of Next-Generation Sequencing Technologies for the Diagnosis of Primary Immunodeficiencies. Front Immunol. 2017; 8: 847. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCasanova JL: Human genetic basis of interindividual variability in the course of infection. Proc Natl Acad Sci U S A. 2015; 112(51): E7118–27. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCasanova JL: Severe infectious diseases of childhood as monogenic inborn errors of immunity. Proc Natl Acad Sci U S A. 2015; 112(51): E7128–37. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAllenspach E, Torgerson TR: Autoimmunity and Primary Immunodeficiency Disorders. J Clin Immunol. 2016; 36(Suppl 1): 57–67. PubMed Abstract \| Publisher Full Text\n\nPicard C, Al-Herz W, Bousfiha A, et al.: Primary Immunodeficiency Diseases: an Update on the Classification from the International Union of Immunological Societies Expert Committee for Primary Immunodeficiency 2015. J Clin Immunol. 2015; 35(8): 696–726. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCasanova JL, Abel L: Human genetics of infectious diseases: Unique insights into immunological redundancy. Semin Immunol. 2018; 36: 1–12. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStray-Pedersen A, Sorte HS, Samarakoon P, et al.: Primary immunodeficiency diseases: Genomic approaches delineate heterogeneous Mendelian disorders. J Allergy Clin Immunol. 2017; 139(1): 232–245. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWolach O, Kuijpers T, Ben-Ari J, et al.: Variable clinical expressivity of STAT3 mutation in hyperimmunoglobulin E syndrome: genetic and clinical studies of six patients. J Clin Immunol. 2014; 34(2): 163–70. PubMed Abstract \| Publisher Full Text\n\nBousfiha A, Jeddane L, Picard C, et al.: The 2017 IUIS Phenotypic Classification for Primary Immunodeficiencies. J Clin Immunol. 2018; 38(1): 129–143. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBerger M, Geng B, Cameron DW, et al.: Primary immune deficiency diseases as unrecognized causes of chronic respiratory disease. Respir Med. 2017; 132: 181–188. PubMed Abstract \| Publisher Full Text\n\nCiancanelli MJ, Abel L, Zhang SY, et al.: Host genetics of severe influenza: from mouse Mx1 to human IRF7. Curr Opin Immunol. 2016; 38: 109–20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCasanova JL, Conley ME, Seligman SJ, et al.: Guidelines for genetic studies in single patients: lessons from primary immunodeficiencies. J Exp Med. 2014; 211(11): 2137–49. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nvon Bernuth H, Picard C, Jin Z, et al.: Pyogenic bacterial infections in humans with MyD88 deficiency. Science. 2008; 321(5889): 691–696. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAlsina L, Israelsson E, Altman MC, et al.: A narrow repertoire of transcriptional modules responsive to pyogenic bacteria is impaired in patients carrying loss-of-function mutations in MYD88 or IRAK4. Nat Immunol. 2014; 15(12): 1134–42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHerman M, Ciancanelli M, Ou YH, et al.: Heterozygous TBK1 mutations impair TLR3 immunity and underlie herpes simplex encephalitis of childhood. J Exp Med. 2012; 209(9): 1567–1582. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSancho-Shimizu V, Pérez de Diego R, Lorenzo L, et al.: Herpes simplex encephalitis in children with autosomal recessive and dominant TRIF deficiency. J Clin Invest. 2011; 121(12): 4889–902. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGuo YQ, Audry M, Ciancanelli M, et al.: Herpes simplex virus encephalitis in a patient with complete TLR3 deficiency: TLR3 is otherwise redundant in protective immunity. J Exp Med. 2011; 208(10): 2083–2098. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCiancanelli MJ, Huang SX, Luthra P, et al.: Infectious disease. Life-threatening influenza and impaired interferon amplification in human IRF7 deficiency. Science. 2015; 348(6233): 448–53. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSpeake C, Presnell S, Domico K, et al.: An interactive web application for the dissemination of human systems immunology data. J Transl Med. 2015; 13: 196. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBerglund LJ, Avery DT, Ma CS, et al.: IL-21 signalling via STAT3 primes human naive B cells to respond to IL-2 to enhance their differentiation into plasmablasts. Blood. 2013; 122(24): 3940–50. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLafaille FG, Pessach IM, Zhang SY, et al.: Impaired intrinsic immunity to HSV-1 in human iPSC-derived TLR3-deficient CNS cells. Nature. 2012; 491(7426): 769–73. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBoisson B, Laplantine E, Prando C, et al.: Immunodeficiency, autoinflammation and amylopectinosis in humans with inherited HOIL-1 and LUBAC deficiency. Nat Immunol. 2012; 13(12): 1178–86. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMorimoto M, Myung C, Beirnes K, et al.: Increased Wnt and Notch signaling: a clue to the renal disease in Schimke immuno-osseous dysplasia? Orphanet J Rare Dis. 2016; 11(1): 149. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDella Mina E, Borghesi A, Zhou H, et al.: Inherited human IRAK-1 deficiency selectively impairs TLR signaling in fibroblasts. Proc Natl Acad Sci U S A. 2017; 114(4): E514–E523. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPark J, Munagala I, Xu H, et al.: Interferon Signature in the Blood in Inflammatory Common Variable Immune Deficiency. PLoS One. 2013; 8(9): e74893. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIsnardi I, Ng YS, Srdanovic I, et al.: IRAK-4- and MyD88-dependent pathways are essential for the removal of developing autoreactive B cells in humans. Immunity. 2008; 29(5): 746–57. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBohn G, Allroth A, Brandes G, et al.: A novel human primary immunodeficiency syndrome caused by deficiency of the endosomal adaptor protein p14. Nat Med. 2007; 13(1): 38–45. PubMed Abstract \| Publisher Full Text\n\nRahman M, Boughorbel S, Presnell S, et al.: A curated transcriptome dataset collection to investigate the functional programming of human hematopoietic cells in early life [version 1; referees: 2 approved]. F1000Res. 2016; 5: 414. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMarr AK, Boughorbel S, Presnell S, et al.: A curated transcriptome dataset collection to investigate the development and differentiation of the human placenta and its associated pathologies [version 2; referees: 2 approved] F1000Res. 2016; 5: 305. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRinchai D, Boughorbel S, Presnell S, et al.: A curated compendium of monocyte transcriptome datasets of relevance to human monocyte immunobiology research [version 2; referees: 2 approved]. F1000Res. 2016; 5: 291. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "45968", "date": "21 Mar 2019", "name": "Bertrand De Meulder", "expertise": [ "Reviewer Expertise Bioinformatics", "transcriptomics" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article presents a tool to gather and make preliminary analyses on a set of curated sequencing datasets related to primary immunodeficiencies.\n\nRegarding replication, I could not reproduce the list of datasets with the criteria that are mentioned in the text. I would have to believe that the authors are indeed reporting their results, and that the manual curation of the dataset list was accurate.\n\nThe tool that is presented has a number of issues:\nThe list of datasets is rather small, and the authors do not mention whether they plan on adding datasets to the list, The tutorials on how to use the tools are only available to registered users, and I could not find how to register Graphical issues are present when displaying the charts, both in-tool and in png images.\n\nIs the rationale for creating the dataset(s) clearly described? Yes\n\nAre the protocols appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and materials provided to allow replication by others? Partly\n\nAre the datasets clearly presented in a useable and accessible format? Partly", "responses": [ { "c_id": "4859", "date": "30 Aug 2019", "name": "Salim Bougarn", "role": "Author Response", "response": "Dear Bertrand, We are thankful to Dr. De Meulder for his positive feed back regarding our manuscript and for the careful revision. We respond to the specific comments and describe the introduced changes to the new version below in the text and in bold. Regarding replication, I could not reproduce the list of datasets with the criteria that are mentioned in the text. I would have to believe that the authors are indeed reporting their results, and that the manual curation of the dataset list was accurate. This is not related to GXB but on the pubmed search criteria. Also, please note that the GEO database is growing over time and for your attention the query was run in date 2017-2-1. I have updated in the paper the total number of datasets found in the GEO database using the criteria mentioned in the paper. The tool that is presented has a number of issues: The list of datasets is rather small, and the authors do not mention whether they plan on adding datasets to the list, The curation of the datasets requires substantial effort so we decide to freeze the addition of dataset for this version. We are planning to update the datasets in a future version of the paper. A sentence was added to the material and method in the paper. The tutorials on how to use the tools are only available to registered users, and I could not find how to register The tutorial is available also for non registered user via the following link: http://pid.gxbsidra.org/dm3/tutorials.gsp. The link in the paper was modified. Graphical issues are present when displaying the charts, both in-tool and in png images. It is not clear what graphical issue exactly you encounter. It is possible to hide info panel information to resize the graphs and get a better visualization. Please note that the GSE29536 datasets was removed from the instance and the datanote (Table 1). The figure 1 was slightly modified. Also, the total number of datasets retrieved, selected and loaded to GXB was modified." } ] }, { "id": "45485", "date": "28 Mar 2019", "name": "John B. Ziegler", "expertise": [ "Reviewer Expertise Clinical immunology including immunogenetics of PIDs." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a useful tool to examine publicly available gene transcript data.\n\nThe paper would be more useful if it included more detailed instructions for its use, perhaps including screenshots to illustrate the text.\n\nThe location of the PubMed link of each sample wasn’t immediately obvious to this reader. I now realise it is accessed by clicking a part of the PubMed.gov icon.\n\nThe meaning of the term “ranking” becomes apparent with use but it would not be burdensome to more experienced readers to have it defined or explained.\n\nThe acronym for induced pluripotent stem cells (iPS) should be defined with first use.\n\nThe sample set “Whole Blood Transcriptional Modules generated on Illumina Hu-6 V2 Beadchips. GSE29536” is listed with the disease category “Immunodeficiencies” but the data presented appears to include none from immunodeficiency settings. I see transcripts from infections, SLE, diabetes, Still’s disease. I can’t get more detail of the study because the PubMed link is wrong. The link provided is https://www.ncbi.nlm.nih.gov/pubmed/24069364 which is actually the link for other listed samples (GSE51404 and GSE51405). I haven’t verified all links in the table.\n\nIf improvements such as those above were addressed I believe the paper would be very useful to those conducting immunological research.\n\nIf improvements such as those above were addressed I believe the paper would be very useful to those conducting immunological research.\n\nIs the rationale for creating the dataset(s) clearly described? Yes\n\nAre the protocols appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and materials provided to allow replication by others? No\n\nAre the datasets clearly presented in a useable and accessible format? Yes", "responses": [ { "c_id": "4860", "date": "30 Aug 2019", "name": "Salim Bougarn", "role": "Author Response", "response": "Dear Dr. Ziegler,We are thankful to Dr. Ziegler for his positive feed back regarding our manuscript and for the careful revision. We respond to the specific comments and describe the introduced changes to the new version below in the text and in bold. The paper would be more useful if it included more detailed instructions for its use, perhaps including screenshots to illustrate the text. We have updated a link to the tutorial (http://pid.gxbsidra.org/dm3/tutorials.gsp) as well as a reference to a paper describing [Speake et. al) on how to use GXB with detailed instructions including screenshots.The location of the PubMed link of each sample wasn’t immediately obvious to this reader. I now realise it is accessed by clicking a part of the PubMed.gov icon. The meaning of the term “ranking” becomes apparent with use but it would not be burdensome to more experienced readers to have it defined or explained.The ranking was explained in material and methods section in the paper The acronym for induced pluripotent stem cells (iPS) should be defined with first use.We have added to the paper the definition of iPS.The sample set “Whole Blood Transcriptional Modules generated on Illumina Hu-6 V2 Beadchips. GSE29536” is listed with the disease category “Immunodeficiencies” but the data presented appears to include none from immunodeficiency settings. I see transcripts from infections, SLE, diabetes, Still’s disease. I can’t get more detail of the study because the PubMed link is wrong. The link provided is https://www.ncbi.nlm.nih.gov/pubmed/24069364 which is actually the link for other listed samples (GSE51404 and GSE51405). I haven’t verified all links in the table.We have removed the GSE29536 from the instance and the datanote and modified the total number of datasets retrieved and loaded into GXB. But also, we modified the table 1 and figure 1.We have checked and verified all PubMed links in other datasets. If improvements such as those above were addressed I believe the paper would be very useful to those conducting immunological research.The requested improvements have been incorporated in the paper. Please note that other minor modification updates were done on the manuscript, such as the total number of datasets found in the GEO database using the criteria mentioned in the paper." } ] } ]	1	https://f1000research.com/articles/8-188
https://f1000research.com/articles/8-1544/v1	30 Aug 19	{ "type": "Opinion Article", "title": "Learning from cases: Analysis of two cases of craniopharyngioma from the 19th to the 21st centuries.", "authors": [ "John R. Apps", "J. Ciaran Hutchinson", "Susan Shelmerdine", "Alex Virasami", "Eduard Winter", "Thomas S. Jacques", "Juan-Pedro Martinez-Barbera", "Owen Arthurs", "Thomas Czech", "J. Ciaran Hutchinson", "Susan Shelmerdine", "Alex Virasami", "Eduard Winter", "Thomas S. Jacques", "Juan-Pedro Martinez-Barbera", "Owen Arthurs", "Thomas Czech" ], "abstract": "This manuscript describes the study of two cases of craniopharyngioma, which have been examined repeatedly over three separate centuries. This includes analysis by Josef Engel in 1839, who sought to uncover the physiological role of the pituitary gland; Jacob Erdheim in 1904, who initially described the disease we now call craniopharyngioma, and recent high resolution MRI and micro-CT imaging and attempted DNA analyses of the tumours. The cases highlight how, rightly or wrongly, our interpretation of data is shaped by the technologies, methodologies and prevailing theories of a given time.", "keywords": [ "Craniopharyngiooma", "historical", "Erdheim", "Engel", "micro-CT" ], "content": "\n\nThe acquisition of knowledge from the study of individual cases is a core component of medical curricula across the world; it has been used to learn about and impart knowledge of human physiology, symptomatology, pathology, and clinical therapy for millennia1 (McLean 2016). Combining cases into series has enabled the definition of diseases, define their natural history and forms an initial level of evidence in the evaluation of therapy2 (Sayre, Toklu et al. 2017). Here, we present two cases of craniopharyngioma, where separate studies over three centuries have impacted on our understanding of the normal function of the pituitary gland and the classification of its pathology.\n\n1828, in the seat of the Austrian Empire, a 33 year old Viennese waiter (Case 1) has been admitted to hospital complaining of longstanding weakness of his arms and legs, an increased need for sleep, headaches, intermittent vomiting, and a progressive deterioration of vision finally leading to complete blindness. Dying in a state of severe emaciation, post mortem examination identified a large cystic lesion in the region of the pituitary which was subsequently retained for further studies.\n\nThis patients tumour was kept, and next studied by Josef Engel, a pathologist completing his PhD thesis, “Über den Hirnanhang und den Trichter“ (On the pituitary and the infundibulum) in 1839, working under the guidance of Carl von Rokitansky, one of the fathers of modern scientific pathology. At this time, the function of the pituitary gland was unknown. Through the characterisation of 12 cases of pituitary pathology, Engel proposed that the pituitary was a primitive version of the “small brain”, now known as the cerebellum. Remarkably, he postulated that the function of the cerebellum was for walking forwards, whereas the pituitary was for walking backwards. A drawing of the sample from the thesis is shown in Figure 1a.\n\nCase 1: Pathological drawing from a) Engel’s thesis in 1839 (Über den Hirnanhang und den Trichter, Medical University Vienna), b) Erdheim’s 1904 paper3, showing a large infra-tentorial cystic lesion (arrow). c) Macroscopic photograph of specimen taken at Great Ormond Street Hospital. scale bar=1cm. d) 3T MRI reconstruction of the tumour matching the macroscopic figure. e) Micro-CT image showing cut-away of the cystic tumour with speckled hyper-intense areas consistent with calcification. P= Pons, M=Medulla, C=Cerebellum. f) Cross-sectional non-contrast micro-CT image showing cyst wall continuous with surrounding structures (arrow).\n\nWhist this seems remarkable today, these conclusions were based on his careful observations. Engel noted several morphological correspondences between the pituitary gland and the cerebellum: both are covered by a tent of dura mater with an opposite semi-circular opening to connect with other parts of the brain, both lie within a bony depression in the midline skull base, both are “kidney-shaped”, both are bordered by a venous ring, the pituitary is surrounded by the arterial circle of Willis, while branches of the vertebral arteries enclose the cerebellum. Thus in some way he considered the pituitary gland a simplified replica of the cerebellum, much as others had considered the relationship between the cerebral hemispheres and the cerebellum. At the same time he also noted important differences; while the pituitary stalk slopes down in an anterior direction towards the gland, the cerebellar peduncle enters the cerebellum in an anterior-posterior direction.\n\nUnder the assumption of analogous form suggesting analogous function, observations regarding the influence of disease states of these structures on movement could be presented in a logical way: with malfunction of the cerebellum leading to contractions of the extensors with a tendency to move backwards, while the opposite is true of advanced stages of pituitary malfunction, namely contraction of the flexors with resulting forward movements, an antagonistic effect on the direction of movement of these two structures could be explained.\n\nA review of Engel’s thesis, published in the Medizinische Jahrbücher des kaiserl. königl. österr. Staates, (Austrian Annual Medical Journal), reflected that “This text is a great testament to the author's honest endeavours and for his talent to work in this field. Even if, over the course of time, the results set out herein, reached through pathological-anatomical examinations and observations, should require modification because of new facts, the methods described and carried out herein as well as the standards of the science still deserve recognition, and in any event, the author has earned lasting credit for sharing such excellent and comprehensive observations”, a testament any PhD student would still be proud to receive today.\n\nAlmost 80 years, two emperors and one unification with Hungary later, the next scientific report of this specimen emerged in the literature in 1904, as part of a large study of pituitary pathology by Jacob Erdheim. In his 200 page paper, “Über Hypophysenganggeschwulste und Hirncholesteatome”, Erdheim defined a cohort of lesions he called “hypophyseal duct tumors”4 (Erdheim 1904). In this paper he describes in detail the pathology of seven such lesions, including that of Josef Ecker. A drawing of the macroscopic pathology of this cases lesion from this paper is shown in Figure 1b. Erdheim hypothesised regarding the existence of two fundamentally different types of pituitary lesions; a “benign” type with a basal papillary morphology and a more aggressive type with histological features resembling those of odontogenic tumours of the jaw, also known as adamantinomas. Following this paper, two of the samples, that described above, and that of a 58-year-old shop keepers widow (Case 2), were stored in Vienna and are currently housed in the Narrentum, part of the Natural History Museum of Vienna.\n\nOver the next century, these tumours were renamed craniopharyngioma, following the influence of the American neurosurgeon Harvey Cushing, who described them as the “most formidable of intracranial tumours”5 (Barkhoudarian and Laws 2013). Craniopharyngiomas are currently classified as either papillary (PCP), predominantly a disease of adults, and adamantinomatous (ACP), the most common tumour of the sellar region in children, based on their histological features and broadly corresponding to the subtypes described by Erdheim6 (Louis, Perry et al. 2016). Advances in molecular profiling have now confirmed differing genetic bases to these two subtypes, with PCP usually harbouring V600E mutations in BRAF and activating mutations in CTNNB1 in ACP, and with differing DNA methylation and gene expression profiles6 (Louis, Perry et al. 2016).\n\nIn a review of Erdheim’s paper in 2015, Case 1, the cystic tumour was re-classified as having features of adamantinomatous pathology7 (Pascual, Rosdolsky et al. 2015). In contrast, case 2, showed features of papillary craniopharyngioma. The macroscopic scale of these samples, along with their good state of preservation, anatomically-oriented display and their well characterised clinical histories provided a unique opportunity to study craniopharyngioma using modern techniques (Figure 1c, Figure 2b).\n\nCase 2. a) Pathological drawing from Erdheim’s 1904 paper, showing a papillary pituitary growth4. b) Macroscopic photograph of the specimen. Note that the specimen has been divided through right temporal lobe. c) 3T MRI image showing close relationship of the tumour (black arrow) to the white matter tracts (white arrow). d) Micro-CT image showing complex structure and relations of the tumour. C=Cerebellum. e) Histological drawing of the tumour by Erdheim. f) Toluidine blue staining of a section of the tumour, showing its papillary epithelial nature.\n\nIn 2016, we published the first histological scale high resolution 3D imaging of tumour invasion using micro-focus computed tomography (micro-CT). By imaging small pieces of human ACP we were able to visualise, at the cellular scale, the invasion of tumour into surrounding tissue, giving insight into the mechanisms of tumour invasion, and the challenges of achieving complete surgical resection8 (Apps, Hutchinson et al. 2016). The samples from these two cases were obtained from Vienna and underwent advanced imaging, initially by 3 Tesla MRI (protocols available on request) then by high resolution micro-CT imaging using a Nikon XT H 225 ST micro-CT scanner, utilising a Molybdenum target to maximise tissue contrast, at 10W scanning power.\n\nImaging successfully enabled visualisation to a resolution of 61.1μm and 84.6μm respectively for the two cases and facilitated detailed 3D reconstructions of the tumours and the surrounding structures (Figure 1d–f, Figure 2c,d) (Videos 1–29,10). Case 1, had speckled hyper-intense foci throughout the cyst wall, consistent with calcification, an imaging feature currently used in making supporting a diagnosis of craniopharyngioma. The cyst, while predominantly separated from local structures, was also focally continuous with the surrounding brain, highlighting the challenges in neurosurgical resection (Figure 1f). In contrast, in case 2, the tumour was more exophytic and heterogeneous in nature. MRI highlighted how the tumour boundary was close to neuronal tracts (Figure 2c).\n\n\n\n1 Data File\n\nVideo 1: Virtual dissection of case 1 by micro-CT imaging. Tumour cyst, with overlying tentorium cerebelli, brain stem, cerebellum and mid-brain. Bright foci indicate calcification within the cyst wall.\n\nhttps://dx.doi.org/10.6084/m9.figshare.9724343.v2\n\n\n\n1 Data file\n\nVideo 2: Virtual dissection of case 2 by micro-CT imaging - Tumour with brain stem, cerebellum, midbrain and part of cerebral hemispheres.\n\nhttps://dx.doi.org/10.6084/m9.figshare.9724703.v3\n\nTissue sections were also taken for histology. Profiles of Case 1’s cyst showed a surrounding fibrous capsule, and Case 2’s tumour showed cellular architecture consistent with a diagnosis of papillary craniopharyngioma (Figure 2f). It is likely that the DNA and protein integrity within the tumours suffered due to inadequate fixation over many decades. Although the tumours had been stored in Formalin (Formaldehyde and water) since Erdheim’s study at the start of the 20th century, Formaldeyhyde only became routinely used in biological practice during the last decade of the 19th century, when Dr Ferdinand Blum discovered its properties as a fixative by accidentally fixing his own fingertips whilst investigating Formaldehyde’s potential antiseptic properties3,11,12 (Blum 1893, Fox, Johnson et al. 1985, Simmons 2014). Thus, unfortunately, the DNA within the samples was not well preserved and all attempts at sequencing the CTNNB1 and BRAF genes in these cases failed.\n\nIn summary, we show two cases which have been investigated over different eras have provided valuable insight into one of the most challenging types of brain tumour. We see how Engel’s interpretation of the function of the pituitary was influenced by the early 19th century theories, and Erdheim’s description of hypophyseal duct tumours developed from thorough detailed pathological examination and how combining patients into case series/cohorts facilitates development in medical understanding of diseases Finally we show how modern, advanced imaging techniques can give remarkable detail as to the macro- and micro- anatomy of tumour growth. Such information is valuable for clinicians treating and researching these tumours, in training those in the field and explaining to patients and relatives the challenges involved in managing these tumours.\n\nThe ability to perform such analyses on these cases is a testament to the foresight of the founding fathers of modern pathology and the law makers in Vienna in establishing the collection, storage and detailed annotation of specimens for scientific study. Whilst the fundamental scientific methodology of detailed observation remains unchanged, the conclusions of the separate studies highlight how our interpretation of data is shaped by the technologies, methodologies and prevailing theories of the time. How these samples will be interpreted in the next century remains to be seen.\n\n\nEthical approval\n\nThis study was performed as part of a larger study, which was approved by a national research ethics committee (REC 13/LO/1494) and all samples handled in accordance with the Human Tissue Act (2004). Routine collection of samples for future analysis was implied under Austrian law at time of collection.\n\n\nData availability\n\nNo data are associated with this article\n\nFigshare: Virtual dissection of 1828 case of craniopharyngioma described by Josef Engel by micro-CT imaging. https://doi.org/10.6084/m9.figshare.9724343.v29\n\nThis project contains the following extended data:\n\nVideo 1.mp4 (Video, Virtual dissection of 1828 case of craniopharyngioma described by Josef Engel by micro-CT imaging)\n\nFigshare: Virtual dissection of 1828 case of craniopharyngioma described by Josef Engel by micro-CT imaging. https://doi.org/10.6084/m9.figshare.9724703.v310\n\nThis project contains the following extended data:\n\nVideo 2.mp4 (Video, Virtual dissection of 1904 case of craniopharyngioma described by Jacob Erdheim by micro-CT imaging)\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).", "appendix": "Grant information\n\nJRA was funded by a Cancer Research UK, Clinical Research Training Fellowship. SCS is supported by a RCUK/UKRI Innovation Fellowship and Medical Research Council (MRC) Clinical Research Training Fellowship [MR/R00218/1]. This award is jointly funded by the Royal College of Radiologists (RCR). NJS is supported by NIHR Senior Investigator award and OJA by NIHR Clinician Scientist fellowship award. NJS receives funding from the Great Ormond Street Hospital Children’s Charity. J.P.M.B. is a Great Ormond Street Hospital for Children’s Charity Principal Investigator. Funding was also received from Children with Cancer UK [15/190]. This work was undertaken at GOSH/ICH, UCLH/UCL who received a proportion of funding from the United Kingdom Department of Health’s NIHR Biomedical Research Centre funding scheme.\n\n\nReferences\n\nMcLean SF: Case-Based Learning and its Application in Medical and Health-Care Fields: A Review of Worldwide Literature. J Med Educ Curric Dev. 2016; 3: pii: JMECD.S20377. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSayre JW, Toklu HZ, Ye F, et al.: Case Reports, Case Series - From Clinical Practice to Evidence-Based Medicine in Graduate Medical Education. Cureus. 2017; 9(8): e1546. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBlum F: Der formaldehyde als härtungsmittel. Zeitschrift der wissenschaft fur mikroskopie. 1893; 10.\n\nErdheim J: Über Hypophysenganggeschwulste und Hirmcholesteatome. Sitzungsb Kais Akad Wissen Math Naturw Klin Padiatr. 1904; 113: 537–726.\n\nBarkhoudarian G, Laws ER: Craniopharyngioma: history. Pituitary. 2013; 16(1): 1–8. PubMed Abstract \| Publisher Full Text\n\nLouis DN, Perry A, Reifenberger G, et al.: The 2016 World Health Organization Classification of Tumors of the Central Nervous System: a summary. Acta Neuropathol. 2016; 131(6): 803–820. PubMed Abstract \| Publisher Full Text\n\nPascual JM, Rosdolsky M, Prieto R, et al.: Jakob Erdheim (1874-1937): father of hypophyseal-duct tumors (craniopharyngiomas). Virchows Arch. 2015; 467(4): 459–469. PubMed Abstract \| Publisher Full Text\n\nApps JR, Hutchinson JC, Arthurs OJ, et al.: Imaging Invasion: Micro-CT imaging of adamantinomatous craniopharyngioma highlights cell type specific spatial relationships of tissue invasion. Acta Neuropathol Commun. 2016; 4(1): 57. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nApps J, Hutchinson, JC, Shelmedine S, et al.: Virtual dissection of 1828 case of craniopharyngioma described by Josef Engel by micro-CT imaging. f1000research.com. Media. 2019. http://www.doi.org/10.6084/m9.figshare.9724343.v2\n\nApps J, Hutchinson, JC, Shelmedine S, et al.: Virtual dissection of 1904 case of craniopharyngioma described by Jacob Erdheim by micro-CT imaging. f1000research.com. 2019. http://www.doi.org/10.6084/m9.figshare.9724703.v3\n\nFox CH, Johnson FB, Whiting J, et al.: Formaldehyde fixation. J Histochem Cytochem. 1985; 33(8): 845–853. PubMed Abstract \| Publisher Full Text\n\nSimmons JEe: Fluid Preservation: A comprehensive reference. 2014; 27–28: 274–275." }	[ { "id": "53215", "date": "02 Sep 2019", "name": "Ashley Grossman", "expertise": [ "Reviewer Expertise Endocrine tumours" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a brief but fascinating study of the original descriptions of craniopharyngiomas as the tumours were able to be re-imaged using contemporaneous techniques.\nI have no major criticisms, but feel that the figures, which are the dominant feature of this manuscript, could be more sharp, possibly at a higher resolution. In addition, in many cases there are arrows in the figures which are not alluded to in the Legends.\nFinally, a minor point, in the text there is an absence of appropriate apostrophes: Page 3 second column, \"This patients's tumour\"; Page 4, first column, \"shop keeper's widow\".\nI also note that the references are given at the end numerically but not in the text.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes", "responses": [] }, { "id": "54329", "date": "27 Jan 2020", "name": "Ruth Prieto", "expertise": [ "Reviewer Expertise craniopharyngioma" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper by Apps JR et al. presents the anatomical, histological and molecular studies of two whole craniopharyngioma-brain specimens from autopsies performed in the 19th and 20th Centuries. They are preserved at the Narrenturm, the site that holds the historical anatomical-pathological collection of Vienna. These tumor specimens, which were instrumental to Viennese pathologist Jakob Erdheim for defining hypophyseal duct tumors or craniopharyngiomas in 1904, are now analyzed with the aid of modern neuroradiological and pathological technologies. Each case corresponds to one of the two major histological craniopharyngioma variants. Case 1 was first macroscopically studied by Joseph Engel, Rokitansky’s disciple and pioneer pathologist, in his original dissertation, “About the Pituitary Gland and the Infundibulum”, published in 1839. This same case was re-examined 65 years later by Jakob Erdheim, who contributed a detailed histological examination of the tumor, which together with 6 additional specimens (including Case 2 in this paper) allowed him to define the new category of hypophyseal duct tumors (hypophysenganggeschwülste). For their part, Apps et al. analyzed both cases with 3T-MRI and micro-CT imaging. Nevertheless, they were unable to demonstrate mutations in β-catenin/CTNNB1 and BRAF genes, the recently identified molecular markers of adamantinomatous and papillary craniopharyngiomas, respectively, due to the poor DNA preservation of the biological tissue of these two old specimens. Apps et al.’s work supports the importance of historical autopsy specimens, such as these forming the collection of Vienna’s anatomical-pathological museum, as a source of valid scientific information now that new technologies are available to further explore complex pathological entities. Moreover, it demonstrates that old individual specimens from museum collections remain useful even over long periods of time. Apps’s study confirms that Erdheim’s thorough histological studies from the beginning of the 20th Century remain completely reliable today. His classification of these two specimens into different craniopharyngioma histological variants, adamantinomatous in Case 1 and squamous-papillary in Case 2, were totally correct. Despite being quite short, the historical background provided by the authors about Engel and Erdheim’s original reports makes reading Apps’s article more pleasant.\nThere are some points that the authors may wish to revise to improve their manuscript:\nThe word “craniopharyngioma” is misspelled in the Keywords section (there are two “o”).\n\nIt is advisable to use “Joseph” (instead of “Josef”), as this was how Dr. Engel spelled his name in his own articles.\n\nAuthors should consider adding the dates of birth and death of the doctors mentioned throughout the manuscript, Joseph Engel (1816-1899), Carl von Rokitansky (1804-1878), Jakob Erdheim (1874-1937).\n\nPage 3 (second column): authors mentioned that Engel studied “12 cases of pituitary pathology”, but it is more accurate to say that he studied “12 cases of tumors of the pituitary gland and/or infundibulum”. I recommend reading and discussing the recent article focused on Engel’s dissertation by Pascual JM et al.1\n\nPage 4 (first column, last paragraph): authors mentioned “Josef Ecker”, which seems to be an incorrect combination of Dr. Engel’s name and that of his patient (Ecker Johann).\n\nThe videos of the micro-CT scans taken from the two cases do not provide the resolution necessary to study the tumor-brain relationships. Could the authors provide higher-quality images? On the other hand, despite the authors also obtaining the MR imaging of these cases, they only provide one axial image of Case 2. It would be advisable to provide mid-sagittal and coronal-trans-infundibular images of both cases, as these sections are the most useful ones to accurately show the anatomical relationships between these craniopharyngiomas and the adjacent vital neurovascular structures, including the hypothalamus and third ventricle.\n\nSpecific comments for Case 1:\nIt might be useful for readers wishing to study the original articles to clarify that Apps’s Case 1 corresponds to both Case 10 in Engel’s dissertation, which analyzed 12 pituitary/infundibulum tumors, and also to Case 5 in Erdehim’s 1904 article that reviewed 7 tumors.\n\nIt is inaccurate to describe Case 1 as a “cystic lesion in the region of the pituitary” (page 3). It would be better described as a cystic tumor at the base of the brain that had replaced the hypophysis and the infundibulum and extended upwards into the third ventricle and downwards into the sphenoid sinus.\n\nFigure 1a: the position of the figure is wrong; it should be positioned upside-down (in order to match the tumor views shown in the remaining panels of the figure). In addition, the arrow should point to the same area as the arrow in figure 1b, pointing to an opening at the cyst wall.\n\nIt would be desirable for the authors to include an image of the histological study performed by Erdheim, as well as their own histological study (as they do in Case 2). On the other hand, since the authors had the chance to take samples of the tumors, if possible, please provide immunostaining for GFAP. On page 5 they stated that the tumor was surrounded by a “fibrous capsule”, but it probably corresponds to reactive gliosis.\n\nSpecific comments for Case 2:\nAs in the previous case, it might be useful to clarify that Case 2 corresponds to Case 6 in Erdehim’s 1904 article.\n\nLegend of Figure 2a: it is not correct to describe this tumor as “a papillar pituitary growth” as Erdheim’s drawing clearly shows that the pituitary gland (“H”) is intact. This tumor would be better defined as “a papillary infundibular growth”.\n\nFigure 2c: I do not understand the reason for the arrow pointing to “white matter tracts”. What really matters is the close relationship between the tumor and the hypothalamus, which is the vital structure to consider when planning surgery of these pathological entities.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1544
https://f1000research.com/articles/8-1531/v1	29 Aug 19	{ "type": "Case Report", "title": "Case Report: Interdisciplinary management of a complex odontoma with a periapical involvement of superior anterior teeth", "authors": [ "Esteban Isaí Flores Orozco", "Amjad Abu Hasna", "Moacir Teotonio de Santos Junior", "Elan Ignacio Flores Orozco", "Renata Falchete Do Prado", "Gabriel Rocha Campos", "Marcia Carneiro Valera", "Esteban Isaí Flores Orozco", "Moacir Teotonio de Santos Junior", "Elan Ignacio Flores Orozco", "Renata Falchete Do Prado", "Gabriel Rocha Campos", "Marcia Carneiro Valera" ], "abstract": "This case report aims to describe the management of a complex odontoma with endodontic involvement of surrounding teeth utilizing a new bioceramic cement consisting of five mineral oxides (5MO) as a retro-filling material. The patient presented for routine consultation with slight dental mobility in the antero-superior region. Upon clinical and computed tomography examination, bone rarefaction was observed in the apical region of teeth 11 and 12, external root resorption in the involved teeth and necrotic pulp of tooth 12. Root canal treatment was performed in teeth 11 and 12. Later, local surgical excision of the lesion was carried out, finding a mass with clinical features of complex odontoma, with histopathological examination of the mass confirming this diagnosis. Retro-filling of tooth 12 with 5MO was carried out. No signs and symptoms were observed over twelve-months of follow-up, with bone neoformation observed in the region. Therefore, 5MO appears to be an effective bioceramic cement that has reparative features.", "keywords": [ "Complex odontoma", "Five mineral oxides", "Bioceramics", "Computed tomography" ], "content": "Introduction\n\nOdontogenic tumors originate from tooth-forming tissues (Siriwardena et al., 2019), and are divided based on their origin into epithelial odontogenic tumors, kerato-cystic odontogenic tumors, mixed odontogenic tumors, and mesenchymal odontogenic tumors (Wright & Vered, 2017).\n\nOdontomas are benign odontogenic tumors of mixed origin (epithelial and mesenchymal), composed mainly of mineralized and conjunctive structure including enamel, dentin, cementum and pulp (Wood & Goaz, 1997). Odontomas are subdivided into compound and complex odontomas; compound odontoma corresponds to small variable structures morphologically similar to teeth, whereas complex odontomas have no anatomical similarity with teeth and present as a structural conglomeration of enamel and dentin (Wright & Vered, 2017).\n\nComplex odontomas are mostly asymptomatic, are usually diagnosed during the first and second decade of life (Budnick, 1976), and are more prevalent in the maxilla than in the mandible (Ahire et al., 2018). They may have a differential radiographic diagnosis with highly calcified lesions like osteoma and adenomatoid odontogenic tumor, among others (Sun et al., 2015). Odontomas are treated by local surgical excision and the prognosis is favorable without recurrence (Gervasoni et al., 2017).\n\nFive mineral oxides (5MO) is a new bioceramic cement that’s use has been described in the literature as an apical plug of open apex teeth (Altaqi, 2014) and a capping material (Ala Rachi et al., 2014). It is a modified MTA (mineral trioxide aggregate) material that consists of five mineral oxides.\n\n\nCase report\n\nA Brazilian white 27-year-old female presented for routine examination with a slight dental mobility in teeth 11 and 12 with no history of trauma or previous treatments. Patient clinical history did not present relevant findings.\n\nThe patient had no symptoms and no history of trauma or previous treatments in this region. Teeth 13 to 23 was tested by pulp vitality test (cold test) carried out by refrigerant gas (Endo Ice, Maquira Dental products industry LTDA–Brazil) and relative isolation using cotton rolls and a dental saliva ejector. The apical region was also tested by digital palpation and perpendicular percussion and was asymptomatic, with no fistula. The depth of its gingival pocket varied between 1–3 mm with various exploring locations and grade I mobility. Panoramic radiographic examination revealed a radiopaque bone-like structure but with a greater density than bone, circumscribed by a thin radiolucent margin in the apical region of teeth 11 and 12 presenting features of complex odontoma (Figure 1). Cone beam computed tomography (CBCT) was indicated to obtain an accurate diagnosis of the lesion and its relationship with the adjacent teeth. CBCT was fundamental in the surgery planning and management in choosing the most appropriate access, which was palatal access in this case.\n\nA) and B) panoramic and periapical radiography showing radiopaque lesion in the upper right incisors region; C) tomographic image of the sagittal section showing the relation of the lesion with the apex of the upper right central and lateral incisors; D) axial reconstruction of odontoma.\n\nFirstly, teeth 11 and 12 underwent endodontic treatment due to the negative response to the pulp vitality test, utilizing a Reciproc system R50/0.05 file (VDW, Munich, Germany) and chlorohexidine gluconate gel 2% (Endogel Essential Pharma, Itapetininga, Brazil) as a lubricator and antimicrobial agent during the preparation of the root canals, and then obturating with gutta-percha and Ah Plus sealer (Dentsply, DeTrey GmbH, Konstanz, Germany). No medication was prescribed before or during the treatment. The patient was advised to take acetaminophen (500 mg, maximum four times a day) in case of pain.\n\nA week later, after tomographic planning, surgical excision of the lesion was carried out under local anesthesia using one anesthetic tube (4% articaine with epinephrine 1: 100,000), with intra-oral access to the mass achieved via intra-sulcular incision of the palatal region from tooth 14 to 24. After detachment of the flap, osteotomy was performed using a surgical carbide drill nº 06 (Angelus Prima Dental Ltda. Londrina / PR, Brazil) under intense irrigation with saline solution.\n\nThe enucleated lesion presented as a hyperdense mass located palatal to the roots of teeth 11 and 12, with clinical features of complex odontoma that involved the apical third of the tooth 11. A bone regularization procedure was carried out, followed by cleaning of the surgical cavity by irrigation with saline solution (Figure 2).\n\nA) Osteotomy; B) total excision of the odontoma; C) repositioning of flap and intra-sulcular suture; D) enucleated complex odontoma.\n\nAfterwards, the hard-dental surfaces were polished with McCall periodontal curette (Fava Metalúrgica São Paulo, SP). Tooth 12 demonstrated external apical resorption; thus, an apicectomy was performed, which was then retro-filled with 5MO (SHAM Dentico, Oman) (Figure 3). No bone graft was indicated. The flap was repositioned, followed by intra-sulcular suturing with 3-0 silk thread (Procare Xuyi Webest Medical Products Co, Jiangsu China). The suture was removed after seven days and the patient progressed well postoperatively without intercurrences.\n\nA) Image of the surgical access after washing and polishing the rigid structures with curette; B) creation of a niche in the apical third for des-obturation; C) placement of 5MO cement in the retro-obturation.\n\nImmediately after sample excision, the mass was immersed in 10% formalin for 48 hours in order to obtain tissue fixation and was sent for histopathological examination under a microscope for differential diagnosis. Routine histological techniques were performed for soft tissues embedding in paraffin. Hard tissues fragments were demineralized in Ethylenediaminetetraacetic acid (EDTA) (TiTriplex® Merk Darmstadt, Germany) for two months, before embedding. The paraffin blocks were cut into sections and slides were subjected to hematoxylin and eosin (HE) staining and observed using an optical microscope. All slides images were obtained using a Pannoramic Scan (3DHISTECH, Budapest, Hungary).\n\nAfter seven days, no postoperative intercurrences were reported by the patient. Through intra oral evaluation, no hematoma or edema was noticed except for the first two days as the patient mentioned. No exudate was noticed.\n\nSoft tissue sections (correspondent to tumor surrounding tissue) demonstrated fragments of connective tissue composed of dense collagen fibers randomly arranged, with fibroblasts and hyalinization areas. Immersed in bundles of collagen fibers, many basophilic round calcification areas were observed. Inflammatory cells were not present. Odontogenic islands were observed.\n\nHistologically, the tumor mass was composed of tooth-like structures exhibiting an irregular arrangement. Hard tissue sections presented fragments of a mixture of material composed mainly of immature dentin, enamel matrix, cementum, and pulp tissue. Dentin immature matrix presented irregular tubules; enamel matrix demonstrated a basophilic pattern and barely prismatic spaces from hydroxyapatite loss during the demineralization process; cementum areas were superimposed to dentin, differing from dentin because of its basophilic irregular lines and the presence of cells inside the matrix (cellular cementum); finally, pulp tissue was observed surrounded by dentin matrix, composed of delicate collagen fibers and fibroblasts. In some places, the dentin matrix surface presented columnar cells similar to odontoblasts. Histopathological examination confirmed the diagnosis of complex odontoma (Figure 4).\n\nA) Connective tissue around the tumor showing bundles of collagen representing mineralized areas (green arrows) and an island of odontogenic epithelium (blue arrow); B) pulpal tissue-like area, demonstrating delicate collagen fibers, a columnar layer of odontoblasts (blue arrows) related to the pre-dentin (yellow arrow) and the dentin matrix showing dentinal tubules (green arrow); C) enamel matrix; D) mineralized tissue similar to cement, presenting basophilic reverse lines (green arrows).\n\nThree follow-up sessions were performed after 30 days, four months and 12 months. In these sessions, clinical intra oral examination was carried out as well as radiographic examination. Bone neoformation was observed by periapical radiograph in the lesion region at the second follow-up session (after four months) compared to the initial one. The periapical radiograph was taken using a radiographic positioner (Figure 5).\n\nA) Periapical radiography of the endodontic treatment before surgery; B) radiographic examination after excision of the odontoma and performing the apical retro-obturation of tooth 12; C) 30 days after surgery; D) and E) 12 months after removal of the odontoma.\n\n\nDiscussion\n\nOdontoma is an anomaly of benign asymptomatic development and slow growth, (Tomizawa et al., 2005). Its early diagnosis enables management of the tumor at the most opportune moment, avoiding further complications and healthy structures being compromised.\n\nOdontoma is more common between the first and second decade of life. Thus, some studies have focused on diagnosis in pediatric patients (Katz, 1989; Tomizawa et al., 2005). In a Brazilian community, de Medeiros et al. (2018) evaluated the frequency of odontogenic tumors over a period of 22 years and verified that odontoma is the most common tumor and is more frequent in the second and third decade of life. This is in agreement with the case described here, as the patient was diagnosed at 27-years-old, without compromising the normal eruption of the permanent teeth. Another similar study conducted over 26 years found odontoma was the 4th most frequently detected tumor in an African community, with greater prevalence in maxilla than in mandible, and in anterior teeth than in premolars and molars (Mamabolo et al., 2011), as found in the present case report.\n\nA further study conducted over 12 years revealed that the prevalence of odontoma is relatively low in African countries in comparison to more developed countries (Aregbesola et al., 2018). Similar results of a low prevalence were found in a Brazilian community, in which odontoma was more frequent in mandible than in maxilla and in females than males (Lima-Verde-Osterne et al., 2017). Furthermore, similar results were found in another study, reporting a low prevalence of odontoma in a 10 year study in an Indian community, with the majority of cases occurring in male patients (Mehngi et al., 2018).\n\nThe use of CBCT was fundamental in this case report; since the indicated treatment of odontoma is the surgical enucleation, CBCT helped in the planning and identification of calcified structures (Sun et al., 2015). CBCT provides a three-dimensional view of the continuity, size and volume of lesions for surgical consideration, at a relatively low radiation dose and high resolution (Favia et al., 2000; Sun et al., 2015).\n\nHistopathological analysis is also important in cases of odontoma. The sample should be carefully examined under the microscope to confirm the differential diagnosis.\n\nMany bioceramics and MTA-like cements induce osteogenic differentiation and stimulate the repair of lesions (Margunato et al., 2015); in this case report, 5MO was effective in promoting apical sealing, which contributed to the osteogenic differentiation and the repair of the lesion. The performed intervention was successful, due to the absence of pathognomonic signs and symptoms, showing a biocompatibility in the presence of 5MO cement in the apical region.\n\nMTA as retro-filling material is better than amalgam and purely gutta-percha in endodontic surgery (Tang et al., 2010). In this case report, the 5MO cement used was effective in a case of periapical involvement and acted as the MTA as it is an MTA-like material.\n\nThe surgical planning using CBCT analysis, the parendodontic intervention through apical retro-filling with the 5MO, and the histopathological analysis of the sample described here reinforces the best pre-established protocol of the treatment (Gervasoni et al., 2017).\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\nWritten informed consent for publication of their clinical details and clinical images was obtained from the patient.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nAhire MS, Tupkari JV, Chettiankandy TJ, et al.: Odontogenic tumors: A 35-year retrospective study of 250 cases in an Indian (Maharashtra) teaching institute. Indian J Cancer. 2018; 55(3): 265–272. PubMed Abstract \| Publisher Full Text\n\nAla Rachi MN, Al-Nahlawi TF, Kouki MT: New Five Minerals Oxides Pulp Capping Material Compared with Dycal. Dent Mater. 2014; 30: e126. Publisher Full Text\n\nAltaqi Z: Apical plug with Five Mineral Oxides (5MO), 3 months follow up. [Online]. 2014; [Accessed: 14 May 2018]. Reference Source\n\nAregbesola B, Soyele O, Effiom O, et al.: Odontogenic tumours in Nigeria: A multicentre study of 582 cases and review of the literature. Med Oral Patol Oral Cir Bucal. 2018; 23(6): e761–e766. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBudnick SD: Compound and complex odontomas. Oral Surg Oral Med Oral Pathol. 1976; 42(4): 501–506. PubMed Abstract \| Publisher Full Text\n\nde Medeiros WK, da Silva LP, Santos PP, et al.: Clinicopathological analysis of odontogenic tumors over 22 years period: Experience of a single center in northeastern Brazil. Med Oral Patol Oral Cir Bucal. 2018; 23(6): e664–e671. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFavia G, Maiorano E, Orsini G, et al.: Central (intraosseous) adenoid cystic carcinoma of the mandible: report of a case with periapical involvement. J Endod. 2000; 26(12): 760–763. PubMed Abstract \| Publisher Full Text\n\nGervasoni C, Tronchet A, Spotti S, et al.: Odontomas: review of the literature and case reports. J Biol Regul Homeost Agents. 2017; 31(2 Suppl 1): 119–125. PubMed Abstract\n\nKatz RW: An analysis of compound and complex odontomas. ASDC J Dent Child. 1989; 56(6): 445–449. PubMed Abstract\n\nLima-Verde-Osterne R, Turatti E, Cordeiro-Teixeira R, et al.: The relative frequency of odontogenic tumors: A study of 376 cases in a Brazilian population. Med Oral Patol Oral Cir Bucal. 2017; 22(2): e193–e200. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMamabolo M, Noffke C, Raubenheimer E: Odontogenic tumours manifesting in the first two decades of life in a rural African population sample: a 26 year retrospective analysis. Dentomaxillofac Radiol. 2011; 40(6): 331–337. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMargunato S, Tasli PN, Aydin S, et al.: In Vitro Evaluation of ProRoot MTA, Biodentine, and MM-MTA on Human Alveolar Bone Marrow Stem Cells in Terms of Biocompatibility and Mineralization. J Endod. 2015; 41(10): 1646–1652. PubMed Abstract \| Publisher Full Text\n\nMehngi R, Rajendra K, Bhagwat P, et al.: Clinical and Histopathological Analysis of Odontogenic Tumors in Institution-A 10 Years Retrospective Study. J Contemp Dent Pract. 2018; 19(10): 1288–1292. PubMed Abstract \| Publisher Full Text\n\nSiriwardena BSMS, Crane H, O’Neill N, et al.: Odontogenic tumors and lesions treated in a single specialist oral and maxillofacial pathology unit in the United Kingdom in 1992-2016. Oral Surg Oral Med Oral Pathol Oral Radiol. 2019; 127(2): 151–166. PubMed Abstract \| Publisher Full Text\n\nSun L, Sun Z, Ma X: Multiple complex odontoma of the maxilla and the mandible. Oral Surg Oral Med Oral Pathol Oral Radiol. 2015; 120(1): e11–6. PubMed Abstract \| Publisher Full Text\n\nTang Y, Li X, Yin S: Outcomes of MTA as root-end filling in endodontic surgery: a systematic review. Quintessence Int. 2010; 41(7): 557–566. PubMed Abstract\n\nTomizawa M, Otsuka Y, Noda T: Clinical observations of odontomas in Japanese children: 39 cases including one recurrent case. Int J Paediatr Dent. 2005; 15(1): 37–43. PubMed Abstract \| Publisher Full Text\n\nWood N, Goaz P: Differential Diagnosis of Oral and Maxillofacial Lesions. 5th ed. Mosby. 1997. Reference Source\n\nWright JM, Vered M: Update from the 4th Edition of the World Health Organization Classification of Head and Neck Tumours: Odontogenic and Maxillofacial Bone Tumors. Head Neck Pathol. 2017; 11(1): 68–77. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "53151", "date": "30 Sep 2019", "name": "Talal Al-Nahlawi", "expertise": [ "Reviewer Expertise Endodonticas and restorative dentistry." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis case report describes a proposal of the use of a new bioceramic cement which is indicated as a retrofitting or a perforation repair material in Endodontics. The abstract is fully comprehensive.\nThe introduction presented a brief explanation of odontoma and the classification of odontoma (complex and compound odontoma) , in addition to a brief phrase the new bioceramic cement of five minerals oxide 5MO. However, more details should be provided about this new cement, such as its composition and its clinical applications.\nThe case report part was well explained, however, more control period is indicated for 2, and 5 years until the full healing is obtained, and make sure about the effectivity of this bioceramic cement.\nThe discussion section was good as well. As a suggestion for the authors, the 5MO cement should be compared in the discussion with the similar bioceramic cements like MTA and its new versions like MTA HP and others, this should be more discussed.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [] }, { "id": "56260", "date": "21 Nov 2019", "name": "Fabio da Silva Matuda", "expertise": [ "Reviewer Expertise Periodontology", "Restorative Dentistry", "Implants" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper is suitable for indexing, but I suggest that was necessary the bioethics committee approval number. The work description is very detailed and the applied exams allow elucidation of the case studied. The pictures are of excellent quality and clearly demonstrate the clinical technique used to remove the lesion.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1531
https://f1000research.com/articles/8-103/v1	25 Jan 19	{ "type": "Case Report", "title": "Case Report: Concurrent primary CNS lymphoma and meningothelial meningioma - nuances of diagnosis and management", "authors": [ "Samir Kashyap", "Jacob Bernstein", "Ira Bowen", "Rosalinda Menoni", "Dan Miulli", "Jacob Bernstein", "Ira Bowen", "Rosalinda Menoni", "Dan Miulli" ], "abstract": "Background: The incidence of two distinct primary intracranial pathologies is an exceedingly rare phenomenon. Although meningiomas are well known to coexist with other primary intracranial malignancies there are only nine reported cases of a meningioma occurring simultaneously with primary CNS lymphoma in the literature. We report a case of a woman who sustained multiple injuries due to two distinct intracranial pathologies, however, lateralizing signs were unrecognized for two weeks prior to her final diagnosis. Case Description: A 64-year-old female with history of diabetes mellitus type 2 initially presented to the Emergency Department, two weeks prior, following a mechanical fall at home resulting in a left bimalleolar fracture. CT imaging revealed a right occipital mass with significant vasogenic edema causing 12mm of midline shift. MRI revealed two distinct homogeneously contrast-enhancing lesions: a right occipital mass with dural-based attachment, as well as a homogenously contrast-enhancing lesion adjacent to the right posterolateral ventricle. FLAIR signal changes were also appreciated and were noted to extend across the corpus callosum, raising concerns for a high-grade glial process. She underwent a right occipital craniotomy with gross total resection of the right occipital mass as well as subtotal resection and biopsy of the second lesion. Final pathology of the extra-axial lesion was found to be meningothelial meningioma and the deep lesion was found to be diffuse large B-cell lymphoma. Discussion: We describe a rare instance of simultaneous meningioma and primary CNS lymphoma that was found to be the underlying cause of a traumatic injury several weeks after the incident. We review the current diagnosis and management nuances in the setting of multiple intracranial oncologic processes.", "keywords": [ "CNS lymphoma", "Meningioma", "Collision tumor" ], "content": "Introduction\n\nThe incidence of two distinct primary intracranial pathologies is an exceedingly rare phenomenon. The reported incidence of such an occurrence is approximately 1 in a million annually (Lee et al., 2002). Although meningiomas, given their benign and slow-growing nature, are well known to coexist with other primary intracranial malignancies such as glioblastoma, metastases, adenomas, there are only nine reported cases of a meningioma occurring simultaneously with primary CNS lymphoma (PCNSL) in the literature (Gordon et al., 2011; Slowik & Jellinger, 1990). Here, we report a case of a woman who sustained multiple injuries due to two distinct intracranial pathologies, however, lateralizing signs were unrecognized for two weeks prior to her final diagnosis.\n\n\nCase presentation\n\nA 64-year-old Hispanic female with a past medical history of type 2 diabetes mellitus and hypertension presented with a chief complaint of left hemiparesis and paresthesias and was activated as a code stroke. History was limited due to the patient being Spanish-speaking only. She did not receive tPA because she stated her left-sided symptoms were not new and she had progressively worsening clumsiness of her left side and that she had been falling to her left. She presented to urgent care two weeks prior to presentation after sustaining a mechanical fall at home. She was diagnosed with a left bimalleolar fracture, placed in a cast, and scheduled for outpatient follow up with orthopedics for surgical evaluation. Computed tomography (CT) of head revealed a right occipital mass with significant vasogenic edema causing 12mm of midline shift (Figure 1).\n\nThe patient was alert and oriented to person, place and time in Spanish. Cranial nerve exam revealed no deficits and no evidence of visual field cut. Motor examination revealed left hemiparesis (4+/5 in the upper and lower extremities), but was limited by the previous casting of her distal left malleolar fracture. Sensory examination revealed slight diminished sensation in the left upper and lower extremities with similar limitations as motor examination.\n\nThe patient was started on dexamethasone 6mg every 6 hours and admitted to the ICU. A STAT MRI brain with and without contrast revealed two homogeneously contrast-enhancing lesions: a 4.8.×6.1×3cm right parieto-occipital extra-axial mass with dural-based attachment, as well as a 3.4×1.8×2.2cm homogenously contrast-enhancing lesion adjacent to the right posterolateral ventricle. FLAIR signal changes were also appreciated and were noted to extend across the splenium of the corpus callosum, raising concerns for a high-grade glial process (Figure 2).\n\nT1-post contrast reveals 2 distinct lesions – a homogenously enhancing extra-axial lesion in the right parietal lobe as well as a homogeneously enhancing periventricular lesion (bottom).\n\nAfter preoperative clearance, a right occipital craniotomy was performed with anticipation for gross total resection of the right parieto-occipital lesion and biopsy with likely subtotal resection and biopsy of the second lesion. Preliminary pathology from intra-operative frozen specimen were consistent with meningioma (extra-axial lesion) and high-grade glioma (periventricular lesion). Gross total resection was performed for the extra-axial lesion and maximal, safe resection of the periventricular lesion was performed. She tolerated the procedure well and had an improved neurological exam postoperatively. Her left hemiparesis improved compared with pre-operative exam, however, she did have very minor left visual field deficits. Post-operative MRI demonstrated gross total resection of meningioma and subtotal resection of what was later confirmed to be diffuse large B-cell lymphoma (Figure 3). During this same admission, she also underwent open reduction, internal fixation (ORIF) of her left bimalleolar fracture without complication. She was discharged home in stable condition.\n\nT1-post contrast reveals gross-total resection of the previously seen extra-axial lesion in the right parieto-occipital region as well as subtotal resection of right periventricular lesion (bottom). The midline shift is significantly improved from pre-operative MRI (Figure 2).\n\nExtra-axial lesion: Meningothelial Meningioma\n\nPeriventricular lesion: Diffuse Large B-Cell Lymphoma (+CD20, +BCL-6, +BCL-2, +MUM-1, +KI67)\n\nAt one month clinic follow-up, she was noted to have an intact motor exam with stable visual field deficits on gross examination. She went into complete remission after a course of methotrexate, cytarabine, and Rituxan and 4 cycles of radiation therapy. She tolerated the treatment relatively well with minor symptoms. At one and two year follow-ups, she continues to be in remission with no signs of recurrence on imaging.\n\n\nDiscussion\n\nWe report a rare case of a concurrent meningioma and primary CNS lymphoma (PCNSL), a rare occurrence entity that has only nine reported cases in the literature. The most common concurrent intracranial tumors reported in the literature are meningioma and glioblastoma (Zhang et al., 2018). It is rare to find two or more primary intracranial tumors simultaneously in patients without previous radiation therapy or underlying phacomatosis such as Neurofibromatosis-2 (NF2). The annual incidence of this phenomenon is estimated to be less than one per million (Gordon et al., 2011; Lee et al., 2002).\n\nAccurate diagnosis is essential as the surgical management of these conditions are opposite of one another. One area in which the management in our patient could be improved is a more accurate history and neurological examination. This was likely affected by the fact that the was a non-English speaker and highlights the importance of accurate history taking with a translator to ensure optimal care. Surgical management of PCSNL is typically limited to biopsy if CSF analysis is inconclusive. This is because PCNSL is particularly chemo- and radiosensitive. Conversely, gross total resection is the gold standard in the management of meningiomas and gliomas (Baraniskin & Schroers, 2014; Gordon et al., 2011; Hoang-Xuan et al., 2015; Korfel & Schlegel, 2013; Muñiz et al., 2014). The same principle applies for steroid administration. The administration of glucocorticoids is not recommended in lymphoma as it could affect the diagnostic yield while it is a mainstay in the treatment of vasogenic edema (Hoang-Xuan et al., 2015). Interestingly in our case, the initiation of high-dose dexamethasone did not affect our diagnosis. The typical diagnostic workup for CNS lymphoma consists of CSF analysis for markers such as IL-10, CXCL13, CD19, CD20 or flow cytometry (Baraniskin et al., 2011; Baraniskin & Schroers, 2014; Muñiz et al., 2014; Rubenstein et al., 2013). Due to the mass effect that is exerted by meningiomas, CSF analysis is difficult without a craniotomy as a lumbar puncture would not be recommended in such a setting. MRI is the gold standard diagnostic modality for meningiomas, however, this is complicated by the fact that CNS lymphoma can mimic any and every intracranial pathology, making it difficult to discern whether lymphoma should be considered as a possibility (Bühring et al., 2001; George et al., 2007; Kulkarni et al., 2012).\n\nThe most common association of two primary intracranial tumors is that of meningioma and glioma (>40 reported cases), however given that these tumors are two of the most common primary intracranial tumors this is thought by many to be coincidental, however associations between the two pathologies have been proposed (Ruiz et al., 2015; Slowik & Jellinger, 1990; Suzuki et al., 2010; Zhang et al., 2018). In a report of two patients with concurrent meningioma and high grade gliomas, Ruiz et al. reported a mutation in K409Q of the KLF4 gene within the meningiomas (Ruiz et al., 2015). Suzuki et al. reported an oncogenic effect due to overexpression of platelet-derived growth factor (PDGF) receptors (Suzuki et al., 2010).\n\nSimultaneous presentations tend to affects adults and have a female predominance due to the nature of meningiomas and their apparent relationship with progesterone and estrogen receptors (Pravdenkova et al., 2006). Since meningiomas typically have an indolent course, this is likely why they are often found concurrently with another primary intracranial pathology. In the setting of simultaneous extra-axial and intra-axial lesions, primary CNS lymphoma must remain a consideration to ensure accurate diagnosis and treatment.\n\n\nConsent\n\nThe patient and her family gave written informed consent for presenting all pertinent clinical information in this case report.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nBaraniskin A, Kuhnhenn J, Schlegel U, et al.: Identification of microRNAs in the cerebrospinal fluid as marker for primary diffuse large B-cell lymphoma of the central nervous system. Blood. 2011; 117(11): 3140–3146. PubMed Abstract \| Publisher Full Text\n\nBaraniskin A, Schroers R: Modern cerebrospinal fluid analyses for the diagnosis of diffuse large B-cell lymphoma of the CNS. CNS Oncol. 2014; 3(1): 77–85. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBühring U, Herrlinger U, Krings T, et al.: MRI features of primary central nervous system lymphomas at presentation. Neurology. 2001; 57(3): 393–396. PubMed Abstract \| Publisher Full Text\n\nGeorge B, Kumar R, Johns P, et al.: Contiguous synchronous occurrence of primary cerebral lymphoma and meningioma. Br J Neurosurg. 2007; 21(1): 35–38. PubMed Abstract \| Publisher Full Text\n\nGordon AS, Fallon KE, Riley KO: Meningioma interdigitated with primary central nervous system B-cell lymphoma: A case report and literature review. Surg Neurol Int. 2011; 2: 181. PubMed Abstract \| Free Full Text\n\nHoang-Xuan K, Bessell E, Bromberg J, et al.: Diagnosis and treatment of primary CNS lymphoma in immunocompetent patients: guidelines from the European Association for Neuro-Oncology. Lancet Oncol. 2015; 16(7): e322–32. PubMed Abstract \| Publisher Full Text\n\nKorfel A, Schlegel U: Diagnosis and treatment of primary CNS lymphoma. Nat Rev Neurol. 2013; 9(6): 317–327. PubMed Abstract \| Publisher Full Text\n\nKulkarni KM, Sternau L, Dubovy SR, et al.: Primary dural lymphoma masquerading as a meningioma. J Neuroophthalmol. 2012; 32(3): 240–242. PubMed Abstract \| Publisher Full Text\n\nLee EJ, Chang CH, Wang LC, et al.: Two primary brain tumors, meningioma and glioblastoma multiforme, in opposite hemispheres of the same patient. J Clin Neurosci. 2002; 9(5): 589–591. PubMed Abstract \| Publisher Full Text\n\nMuñiz C, Martín-Martín L, López A, et al.: Contribution of cerebrospinal fluid sCD19 levels to the detection of CNS lymphoma and its impact on disease outcome. Blood. 2014; 123(12): 1864–1869. PubMed Abstract \| Publisher Full Text\n\nPravdenkova S, Al-Mefty O, Sawyer J, et al.: Progesterone and estrogen receptors: opposing prognostic indicators in meningiomas. J Neurosurg. 2006; 105(2): 163–173. PubMed Abstract \| Publisher Full Text\n\nRubenstein JL, Wong VS, Kadoch C, et al.: CXCL13 plus interleukin 10 is highly specific for the diagnosis of CNS lymphoma. Blood. 2013; 121(23): 4740–4748. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRuiz J, Capilla E, Díaz JF, et al.: Secretory meningioma with KLF4 K409Q mutation in collision with glioma. Clin Neuropathol. 2015; 34(6): 322–329. PubMed Abstract \| Publisher Full Text\n\nSlowik F, Jellinger K: Association of primary cerebral lymphoma with meningioma: report of two cases. Clin Neuropathol. 1990; 9(2): 69–73. PubMed Abstract\n\nSuzuki K, Momota H, Tonooka A, et al.: Glioblastoma simultaneously present with adjacent meningioma: case report and review of the literature. J Neurooncol. 2010; 99(1): 147–153. PubMed Abstract \| Publisher Full Text\n\nZhang Z, Yang Y, Zhang K, et al.: Collision Tumor of Glioblastoma and Meningioma: Case Report and Literature Review. World Neurosurg. 2018; 117: 137–141. PubMed Abstract \| Publisher Full Text" }	[ { "id": "49515", "date": "17 Jun 2019", "name": "Eric Bessell", "expertise": [ "Reviewer Expertise International expert in primary CNS lymphoma and co-author of European Guidelines in 2015." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors are correct in stating that a concurrent presentation of meningioma and primary CNS lymphoma (PCNSL) is very rare and worth publishing. The authors imply that the use of translators for the hispanic population in California is not routine. They should clarify this for an international audience. PCNSL can be rapidly progressive and urgent investigation is essential. Most cases of PCNSL are diagnosed by stereotactic biopsy +/- CSF cytology and a surgical approach is uncommon. In this case the radiological appearance was of a meningioma as well as a lesion adjacent to the right postero-lateral ventricle. The frozen section of the latter lesion suggested a high-grade glioma and for this reason surgical removal of both tumours was indicated. No detailed information is given on the staging investigations carried out which should include CT or PET/CT scanning, CSF cytology, bone marrow aspirate and trephine, slit-lamp examination of the eye and blood tests including LDH. Also baseline neuro-cognitive tests are indicated at the age of 64 years. No detailed information is given on the chemotherapy given, the regimen with methotrexate, cytarabine and rituximab is not referenced and no doses are given and the number of cycles not stated. In addition the radiotherapy dose is in Gy with the number of fractions stated not in cycles. This information is essential in a patient of this age. The authors state that at 2 years there is no recurrence but again give no indication of whether there has been any cognitive deterioration. In the discussion the authors discuss the other brain tumours that occur concurrently with meningioma but don't discuss the outcome achieved in the other 9 cases reported and how they compare with the case they are reporting. In summary this is an interesting case report which would be greatly enhanced if the additional information I have alluded to was included.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Partly\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Partly", "responses": [] }, { "id": "51975", "date": "20 Aug 2019", "name": "Kuntal Kanti Das", "expertise": [ "Reviewer Expertise Neurooncology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors report an interesting case of co-occurrence of primary CNS lymphoma and meningioma in a 64-year-old lady and discuss the importance of preoperative suspicion in their management. The case is indeed interesting and clearly, the authors managed the case well.\n\nMy observations which the authors may like to clarify include:\nWas the presence of hemiparesis not suspected when the patient presented with a fall 2 weeks earlier? Clearly, the tendency to fall on the left side must have been given due importance.\n\nDetails of the surgical approach are necessary. How was the periventricular lesion approached? Through a separate corticectomy or through the bed of the meningioma cavity? Was any intraoperative image guidance utilized?\n\nDid you try to rule out systemic lymphoma? If so, how?\n\nApart from being a function of increased age, is there any connecting link between lymphoma and meningioma? Some more literature needs to be incorporated in this regard.\n\nApart from reporting a rare association, a case like this can be more educative if newer insights can be provided as to how do they co-occur and how should they be managed. I would be happy if the authors address these minor points.\n\nIs the background of the case’s history and progression described in sufficient detail? Partly\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Partly\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-103
https://f1000research.com/articles/8-1518/v1	28 Aug 19	{ "type": "Research Article", "title": "Performance characteristics of a modified HIV-1 drug resistance genotyping method for use in resource-limited settings", "authors": [ "Edwin O. Magomere", "Donald D. Nyangahu", "Sammy Kimoloi", "Brenda A. Webala", "Bartholomew N. Ondigo", "Donald D. Nyangahu", "Sammy Kimoloi", "Brenda A. Webala", "Bartholomew N. Ondigo" ], "abstract": "Background: HIV-1 drug resistance (HIVDR) assays are critical components of HIV clinical management programs in the face of emerging drug resistance. However, the high costs associated with existing commercial HIVDR assays prohibit their routine usage in resource-limited settings. We present the performance characteristics of a modified commercial HIVDR testing assay. Methods: A total of 26 plasma samples were used to validate and assess the accuracy, precision, reproducibility and amplification sensitivity of a modified HIVDR assay by HIV genotyping. In addition, a cost comparison between the original and the modified assay was performed using the ingredient costing approach. Results: The performance characteristics of the modified assay were in agreement with the original assay. Accuracy, precision and reproducibility showed nucleotide sequence identity of 98.5% (confidence interval (CI), 97.9–99.1%), 98.67% (CI, 98.1–99.23) and 98.7% (CI, 98.1–99.3), respectively. There was no difference in the type of mutations detected by the two assays (χ2 = 2.36, p = 0.26). Precision and reproducibility showed significant mutation agreement between replicates (kappa = 0.79 and 0.78), respectively (p < 0.05). The amplification sensitivity of the modified assay was 100% and 62.5% for viremia ≥1000 copies/ml and <1000 copies/ml respectively. Our assay modification translates to a 39.2% reduction in the cost of reagents. Conclusions: Our findings underscore the potential of modifying commercially available HIVDR testing assays into cost-effective, yet accurate assays for use in resource-limited settings.", "keywords": [ "HIV-1 drug resistance testing", "Assay validation", "Accuracy", "Precision", "Reproducibility", "Amplification sensitivity" ], "content": "Introduction\n\nThere has been an unprecedented increase in support for HIV diagnosis, treatment and monitoring programmes. Various multinational groups including the U.S. President’s Emergency Plan for AIDS Relief (PEPFAR) and Global Fund have increased funding of HIV management programs leading to improved access to antiretroviral therapy (ART)1,2. The outcome of the expanded access to ART is significant decline of HIV/AIDS-associated morbidity and mortality2. As a result, there has been an increase in both life expectancy and duration for patients on lifelong ART translating to high risk of HIV drug resistance (HIVDR) development. To sustain the success achieved following improved ART coverage, continuous clinical, virologic and immunological monitoring of patients on ART is required to ensure these positive treatment outcomes are maintained3,4. Unfortunately, development and transmission of HIV drug resistance threatens to derail these achievements5. HIV drug resistance testing (DRT) is routinely used for clinical care in high-income countries; however, it is not available to a majority of patients in resource-limited settings due to the high costs of implementation and limited trained manpower.\n\nThe World Health Organization (WHO) launched the Global Action Plan (GAP) against HIVDR for resource-limited countries whereby two of the proposed approaches included continuous innovation and capacity building of laboratory staff6. Scientists have attempted to develop alternative affordable methods for HIV drug resistance monitoring to improve access5,7–9. One such test was developed by Centre for Disease Control and Prevention (CDC) and is currently distributed by Thermo Fisher Scientific10, and is herein referred to as the original assay. The assay is used to genotype the genetically diverse HIV-1 virus from plasma samples to detect resistance mutations in the protease and reverse transcriptase genes with a good subtype inclusivity rate compared to other commercial DRT assays in the market11.\n\nIn the present study, we modified this commercial HIV DRT protocol by a 50% reduction of reagent volumes for nested PCR and cycle sequencing reactions for genotyping of HIV-1 drug resistance. The performance characteristics of the modified assay were assessed and evaluated for suitability and reliability using WHO guidelines for assay validation. In addition, cost comparison was performed to determine the cost implication of the assay modification.\n\n\nMethods\n\nA total of 26 blood samples were collected in EDTA tubes from patients attending Kenyatta National Hospital HIV clinic laboratory for routine viral load monitoring. Plasma was separated by centrifugation at 2000g for 10 minutes. The plasma samples were stored at -80°C freezer (Nuve, Ankara, Turkey). Remnant plasma samples were used for this study.\n\nEthical clearance for the study was obtained from University of Nairobi-Kenyatta National Hospital Ethics Review Committee (UoN-KNH-ERC). The need for consent was waived since remnant plasma from HIV viral load testing were used in this study.\n\nHIV-1 RNA was extracted from 500 µl plasma using the PureLinkTM extraction kit according to manufacturer’s instructions (Invitrogen, Carlsbad, CA, USA). Briefly, 25 μl proteinase K and 500 μl plasma were transferred into 2-ml microcentrifuge tube. Lysis buffer (500 μl) was added to the mixture, vortexed for 15 seconds, and incubated at 56°C for 15 minutes. Next, 500 μl of 96% ethanol was added to the reaction tube, and vortexed for 15 seconds and incubated at room temperature for 5 minutes. The lysate was transferred to sterile viral spin column, washed twice with 500 μl of wash buffer and finally eluted in 40 μl of elution buffer. RNA was stored at -80°C. The extraction procedure was similar for both the original and modified method.\n\nOriginal genotyping method. The Thermo Fisher ScientificTM HIV Genotyping workflow that amplifies a 1.1-kb fragment covering codons 6–99 of the protease gene and codons 1–251 of the reverse transcriptase (RT) gene was implemented according to manufacturer’s instructions10. Briefly, reverse transcriptase PCR and nested PCR were performed using the HIV-1 Genotyping Kit: Amplification Module. Primers used are shown in Table 1. Cycle sequencing was performed using the HIV-1 Genotyping Kit: Cycle Sequencing Module. The resulting cycle sequencing products were analyzed using an ABI 3730 genetic analyzer. The consensus sequences were generated using ReCall (requires free registration) and drug resistance mutations were interpreted using the Stanford HIVdb genotyping resistance interpretation algorithm.\n\nModified genotyping system Specific reaction steps of the original protocol were modified by reducing the reagent volumes by half.\n\nThe HIV Genotyping kit: Amplification module was used for RT-PCR according to the manufacturer’s instructions. Briefly, 10 µl RNA was denatured by incubating at 65°C for 10 minutes and added to 40 µl RT-PCR Master Mix containing SuperScriptTM III One-Step RT-PCR with PlatinumTMTaq High Fidelity Enzyme. The reaction conditions for RT-PCR were 50°C for 45 minutes where first-strand cDNA synthesis was performed. Enzyme inactivation and denaturation of cDNA-RNA hybrid was accomplished by incubating the reaction at 94°C for 2 minutes. Second-strand synthesis and PCR amplification was carried out in 40 cycles of 94°C for 15 seconds, 50°C for 20 seconds, 72°C for 2 minutes and a final extension for 10 minutes at 72°C using Veriti thermocycler.\n\nNested PCR reaction mix was prepared by adding 2 μl RT-PCR product to 23.7 μl of nested mastermix containing 0.12 mM of each of inner primers, 1X GeneAmp Gold Buffer II, 2 mM MgCl2, 400 mM each dNTP and 0.25 μl AmpliTaq Gold™ LD DNA Polymerase enzyme (Applied Biosystem, CA, USA). The target pol region was amplified using Veriti Thermocycler in a final reaction volume of 25.7 µl (Applied Biosystems, CA, USA). The reaction conditions were 940C for 4 minutes, 40 cycles of 94°C for 15 seconds, 55°C for 20 seconds, 72°C for 2 minutes and a final extension at 72°C for 10 minutes. Nested PCR products of 1.1 kb were confirmed by gel electrophoresis. A 5 μl aliquot of nested PCR product was added to 4 μl ExoSAP-ITTM PCR Purification Reagent) and incubated at 37°C for 15 min and 80°C for 15 minutes in a Veriti thermocycler.\n\nCycle sequencing was performed using the HIV Genotyping kit: Cycle Sequencing module. Six overlapping primers, labelled F1, F2, F3, R1, R2, and R3 were used. One microliter nested PCR product was added to 9 µl of each cycle sequencing mix. The cycle sequencing reaction conditions were 25 cycles of 96°C for 10 seconds, 50°C for 5 seconds, and 60°C for 4 minutes.\n\nThe Big Dye XTerminator purification kit was used to purify the sequencing reaction by adding 10 µl of the Big Dye XTerminator and 45 µl SAM solution to cycle sequencing products. The reaction plate was vortexed at 1,800 rpm for 30 minutes. The plate was then centrifuged at 1000g for 2 minutes at room temperature. Next, 30 μl of purified cycle sequencing products were transferred to a reaction plate and analyzed using an ABI 3730 genetic analyzer (Applied Biosystem, CA, USA). Sequences were generated using ReCall and drug resistance mutations interpreted using the Stanford HIVdb genotyping resistance interpretation algorithm (output files available as Extended data12) and the International AIDS Society (IAS) 2011 mutation list13.\n\nPerformance characteristics of the modified assay were assessed using the WHO/HIV ResNet guidelines, including accuracy, precision, reproducibility and amplification sensitivity14.\n\nAccuracy. Accuracy refers to the agreement between a result and an expected reference value. In genotyping assays, nucleotide sequence identity is used. Ten samples were analyzed using both methods and the degree of concordance in mutations identified was compared based on the 2017 IAS mutations list13. Nucleotide sequence identity between the paired sequences was assessed using the EMBOSS pairwise alignment tool. The WHO recommends 90% nucleotide sequence similarity as the cutoff point for assay performance characteristics14.\n\nPrecision. This is the ability of an assay to generate the same result on multiple replicates of the same sample within a test run. Three samples were analyzed using the modified method in n=4 replicates. The degree of concordance of detected mutations within replicates was determined. Nucleotide sequence identity was also determined using the EMBOSS program for pairwise alignment14.\n\nReproducibility. This is the ability of a test to produce the same result on multiple aliquots of the same sample in different test runs. Ten samples were analyzed in duplicates using the modified method on different days. Nucleotide sequence identity of sequences obtained was assessed using EMBOSS program for pairwise alignment14.\n\nAmplification sensitivity. Amplification sensitivity is defined as the percentage of successful genotyping tests amongst specimens with a specific viral load range. In this study, sixteen samples with viral loads ranging between 207 and 86,040 copies/ml were analyzed using the modified assay to determine the viral load ranges at which ≥95% of the samples were successfully genotyped14.\n\nAn ingredient costing approach was utilized to estimate reagent costs for both original and modified assays at RNA extraction, DNA/RNA amplification, gel electrophoresis, and sequencing steps. All costs were converted to US dollar using 2019 conversion rate.\n\nQuantitative variables were expressed as mean ± standard deviations (SD). The McNemar test was used to assess significance in the discordant mutations between the modified and the original assay. Precision and reproducibility were assessed using the Cohen kappa statistic. Wilcoxon signed-rank test was used to compare the original and modified assays in base calling for mixed bases between the two methods. Statistical Package for Social Sciences (SPSS) version 22 (IBM, NY, USA) was used for all data analysis.\n\n\nResults\n\nAccuracy of sequence identity and mutation detection. The mean nucleotide identity was 98.5% (confidence interval (CI), 97.92–99.1%). A total of 68 mutations were detected, including 9 (13%) mutations in the protease gene and 59 (87%) in the reverse transcriptase (RT) gene as shown in Table 2. The original assay detected 67 drug resistance mutations, including 2 mutations (V106I and K225H) which were missed by the modified assay. On the other hand, the modified assay detected 68 drug resistance mutations, including four mutations (D67DG, K101KPQT, K70KN and I50IL) which were not detected by the original method. A total of six mutations were discordant (V106I, P225H, D67DN, I50IL, K101KPQT, and K70KN). Of the six discordant drug resistance mutations, two were completely discordant (V106I and P225H) while four (D67DN, K101KPQT, K70KN and I50IL) were partially discordant. Three of the discordant mutations (D67DN, K101KPQT, and K70KN) were detected as mixtures in the modified assay as non-mixtures (D67N, K101P and K70N) in the original assay and as shown in Table 3. A high mutation concordance was obtained between the two assays at χ2 = 2.36, p = 0.26. The significance of mixture base-calling between the two methods was determined using the Wilcoxon signed rank test, which showed no significant difference in mixtures detected by the two methods at p = 0.089.\n\nᵃDiscordant mutations detected by the original assay.\n\nᵇDiscordant mutations detected by the modified assay.\n\nᵃDiscordant drug resistance associated mutations detected in original assay.\n\nᵇDiscordant drug resistance associated mutations detected in the modified assay.\n\nPrecision and reproducibility of the modified assay. Assessment to examine the precision (intra-assay precision) and reproducibility (inter-assay precision) of the modified assay were performed by analyzing n=3 samples in quadruplicate in a single test run for precision. The mean nucleotide sequence identity within the replicates was 98.67% (CI, 98.1–99.3). Reproducibility was assessed by testing 10 samples in duplicate on different days using the modified assay. The mean nucleotide sequence identity was 98.6% (CI, 98.2–99.0). The overall agreement of drug resistance mutations detected by precision and reproducibility was significant at kappa value of 0.792 and 0.778, respectively, as shown in Table 4.\n\nAmplification sensitivity of the modified assay. A total of 16 samples with a median viral load of 832.5 copies/ml (IQR 403.5–2454.25) were used to assess amplification sensitivity. The modified assay showed an amplification sensitivity of 100% for samples with viral load ≥1000 copies/ml and 62.5% for samples with viral load <1000 copies/ml as shown in Table 5.\n\nᵇ Failed to amplify\n\nAfter a 50% reduction in reagent volumes in the amplification and sequencing reactions, we compared the cost between the original and the modified method. The cost of analyzing one sample from extraction to sequencing was decreased from $97 to $59. This represents about 39.2% cost reduction as shown in Table 6.\n\n\nDiscussion\n\nHere, we show that the results of the modified assay are comparable to those produced by the original assay in the performance characteristics analyzed including accuracy, precision, reproducibility and amplification sensitivity. We observed a high concordance of drug resistance mutations detected by both original and modified assays. When testing for accuracy, we observed high nucleotide sequence similarity (98.5±0.94%) versus original assay. In addition, we reported 98.67% (CI, 98.1–99.3) sequence similarity when assessing precision and 98.6% (CI, 98.2–99.0) when assessing reproducibility of the modified assay15. Furthermore, the modified assay showed 100% sensitivity for VL >1000 copies/ml which is the recommended indicator for HIV treatment failure16. In general, the modified assay met all the requirements in the WHO Genotyping assay validation recommendations. Altogether our findings demonstrate that reducing the volumes of reagents in the original assay by up to half results in a low-cost HIV drug resistance assay that could be utilized in resource-limited settings without compromising the quality of testing.\n\nAccuracy was assessed by analyzing 10 samples with viral load ranges (4258 –78924 copies per ml) by both the original and the modified assays and all 10 samples were successfully amplified and genotyped with high nucleotide sequence identity (98.5±0.94%). All the clinically relevant mutations were reproducible by both methods in spite of some minor discordance between the two methods. The WHO recommends a minimum of 90% sequence similarity for genotyping assays14. Our findings of 98.5±0.94% nucleotide sequence similarity are in agreement with those reported by Zhou et al.7 who reported 99.41%. In addition, analysis of mutation concordance between the original and the modified assay showed high mutation concordance also reported by Inzaule et al.8. Our findings for mutation concordance (93%) were lower than reported by Manasa et al. (100%)5. The modified assay precision assessment showed a high degree of nucleotide sequence identity within replicates. Similar results for precision (98.22%) were reported by Zhou et al.7. High sequence similarity was also obtained for reproducibility of the modified assay. Our findings for reproducibility (98.6%) were similar to those reported by Zhou et al., who obtained (98.94%)7. Amplification sensitivity was defined as the percentage of samples successfully genotyped at a given plasma viral load range. In this study, all samples with viral load ≥1000 copies/ml were successfully genotyped, a finding that was similar to that reported by a previous study8. For samples with plasma viral loads <1000 copies/ml, our modified assay successfully genotyped 62.5% of the samples which is slightly below 63.6% reported by a previous study8. The WHO criteria for HIVDR testing requires that patients with persistent viral load ≥1000 copies/ml to be tested for HIV drug resistance17. Our assay was 100% efficient in genotyping these samples. Furthermore, with 62.5% amplification sensitivity for low level viremia (LLV) samples, highlighted the possibility of using the modified assay in genotyping patients with LLV. However, there is need to perform further large-scale studies to clearly establish the performance of the modified method in LLV patients.\n\nWe detected six discordant mutations in both original and modified assays. Out of the six discordant mutations, four were mixed-base mutations, which were detected exclusively by the modified assay. The differential detection of discordant mutations between the two assays could be attributed to detection of nucleotide mixtures as a result of subjectivity in basecalling or amplification bias18,19. Previous studies have also reported a significant contribution of nucleotide mixtures to discrepancies of mutations between Viroseq and an in-house assay6,7.\n\nThe comparison of cost between the two assays revealed a reduction of HIV drug resistance testing cost by 39.2% per test. These findings suggest that access to HIV drug resistance services in resource-limited settings could be increased through innovative modifications of existing commercially available assays. A small sample size was used because this was a small modification of a validated test already in use. The applicability of this assay can be demonstrated further by testing a larger number of samples. One limitation of the study is the samples used were from patients failing first line treatment and not exposed to protease inhibitors leading to under representation of drug resistant mutations in the protease gene.\n\n\nConclusion\n\nOur findings underscore the potential utility of a modified cost-effective HIV-1 drug resistance assay for testing plasma samples in resource-limited settings. The performance characteristics of the modified assay were satisfactory and therefore the cheaper assay could be one of the approaches of increasing access to HIVDR tests.\n\n\nData availability\n\nGenBank accession numbers for HIV-1 pol and gag sequences isolated from each participant using each technique are available in Table 7.\n\nFigshare: Performance Characteristics of a Modified HIV-1 Drug Resistance Genotyping Method for use in Resource Limited Settings: https://doi.org/10.6084/m9.figshare.911483312.\n\nThis study contains the following extended data:\n\nAccuracy testing data (Stanford HIV Drug Resistance reports; PDF)\n\nPrecision testing data (Stanford HIV Drug Resistance reports; PDF)\n\nReproducibility testing data (Stanford HIV Drug Resistance reports; PDF)\n\nEM_Accuracy_Precision_Repro_AmpSensitivity (summary of HIV Drug Resistance Mutations detected; xlsx)\n\nGel images (TIF)\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nThe authors thank the study participants and their families.\n\n\nReferences\n\nGlobal Fund: The U . S . & The Global Fund to Fight AIDS, Tuberculosis and Malaria [Report]. 2019. Reference Source\n\nWHO and UNAIDS: Towards Universal Access HIV/AIDS Global Launch of the 2010 Report. [Report]. 2010. Reference Source\n\nBertognolio S, Derdelinckx I, Parker M, et al.: World Health Organization/HIVResNet Drug Resistance Laboratory Strategy. Antivir Ther. 2008; 13 Suppl 2: 49–57. PubMed Abstract\n\nde Waal R, Lessells R, Hauser A, et al.: HIV drug resistance in sub-Saharan Africa: public health questions and the potential role of real-world data and mathematical modelling. J virus Erad. 2018; 4(Suppl 2): 55–58. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nManasa J, Danaviah S, Pillay S, et al.: An affordable HIV-1 drug resistance monitoring method for resource limited settings. J Vis Exp. 2014; (85): 1–13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWHO Global Action Plan: Global action plan on HIV drug resistance 2017-2021. [Report]. 2017. Reference Source\n\nZhou Z, Wagar N, DeVos JR, et al.: Optimization of a low cost and broadly sensitive genotyping assay for HIV-1 drug resistance surveillance and monitoring in resource-limited settings. PLoS One. 2011; 6(11): e28184. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nInzaule S, Yang C, Kasembeli A, et al.: Field evaluation of a broadly sensitive HIV-1 in-house genotyping assay for use with both plasma and dried blood spot specimens in a resource-limited country. J Clin Microbiol. 2013; 51(2): 529–39. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChaturbhuj DN, Nirmalkar AP, Paranjape RS, et al.: Evaluation of a cost effective in-house method for HIV-1 drug resistance genotyping using plasma samples. PLoS One. 2014; 9(2): e87441. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThermoFisher: HIV-1 Genotyping Workflow. [Kit insert], Applied Biosystems cat no. A32317. 2016. Reference Source\n\nCenter of Disease Control: One step closer to realizing the 90-90-90 target. [Report], 2016. Reference Source\n\nMagomere E, Nyangahu D, Kimoloi S, et al.: Performance Characteristics of a Modified HIV-1 Drug Resistance Genotyping Method for use in Resource Limited Settings. figshare. Dataset. 2019. http://www.doi.org/10.6084/m9.figshare.9114833\n\nWensing AM, Calvez V, Günthard HF, et al.: 2017 Update of the Drug Resistance Mutations in HIV-1. Top Antivir Med. 2017; 24(4): 132–41. PubMed Abstract \| Free Full Text\n\nWHO: Recommended Methods for Validation of an In-House Genotyping Assay for Surveillance of HIV Drug Resistance. 2018; 1–15.\n\nChen JH, Wong KH, Li PC, et al.: In-house human immunodeficiency virus-1 genotype resistance testing to determine highly active antiretroviral therapy resistance mutations in Hong Kong. Hong Kong Med J. 2012; 18(1): 20–4. PubMed Abstract\n\nUSA Department of Health and Human Service: Guidelines for the use of antiretroviral agents in adults and adolescents living with HIV. Department of Health and Human Services Washington, DC; 2018. Reference Source\n\nWHO: Guidelines for managing advanced HIV disease and rapid initiation of antiretroviral therapy. [Report], Geneva: World Health Organization. 2017. Reference Source\n\nGalli RA, Sattha B, Wynhoven B, et al.: Sources and magnitude of intralaboratory variability in a sequence-based genotypic assay for human immunodeficiency virus type 1 drug resistance. J Clin Microbiol. 2003; 41(7): 2900–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nParkin N, Bremer J, Bertagnolio S: Genotyping external quality assurance in the World Health Organization HIV drug resistance laboratory network during 2007-2010. Clin Infect Dis. 2012; 54 Suppl 4: S266–72. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "53075", "date": "18 Sep 2019", "name": "Michael Walekhwa", "expertise": [ "Reviewer Expertise Microbial and vaccine immunology" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nTitle\nOk.\nAbstract\n‘We present the performance characteristics of a modified commercial HIVDR testing assay.’ Should be changed to ‘this study established the performance characteristics of a modified commercial HIVDR testing assay’.\n\nThe abstract lacks a recommendation.\n\nThe key word ‘assay validation’ is missing in the abstract.\nIntroduction\n‘Various multinational groups’ use ‘agencies’ instead of ‘groups’.\n\n‘….leading to improved access to antiretroviral therapy (ART)’ replace with ‘leading to improved access to HIV clinical care’. (HIV care is not limited to antiretroviral therapy only).\n\n‘….the outcome of the expanded access to ART is significant decline of HIV/AIDS associated morbidity and mortality.’ Replace with ‘…the outcome of the expanded access to clinical care has been a significant drop in HIV/AIDS related morbidity and mortality.’\n\n‘As a result, there has been an increase in both life expectancy and duration for patients on lifelong ART translating to high risk of HIV drug resistance (HIVDR) development’. This statement has not been brought out clearly & what aspect of ‘duration’ is being implied in this case?\n\n‘……transmission of HIV drug resistance’ add ‘strains’ after resistance.\n\n‘One such test was developed by Centre for Disease Control and Prevention (CDC) and is currently distributed by Thermo Fisher Scientific10, and is herein referred to as the original assay’. Change to ‘One such test, herein referred to as the original assay, was developed by the Centre for Disease Control and Prevention (CDC) and is currently distributed by Thermo Fisher Scientific.’\nMethods\nSamples\nAre 26 samples enough to validate an assay? And how did you arrive upon this sample size?\n\nHow were the samples collected? By venipuncture or cardiac puncture? Be very clear by which method.\n\nAfter clinic, delete the word ‘laboratory’. Patients attend clinics, not clinic laboratories.\n\n-80 is already a freezing condition. Delete the word ‘freezer’, the statement will still make sense without the subtle repetition.\n\nWhat are ‘remnant plasma samples’? Were they the ones stored in the freezer? What defines them as ‘remnant’?\nEthical approval\n\nIt is claimed here that consent was waived, yet in the sample collection, it is clearly stated that the samples were collected from patients. How and why was the consent waived, and by whom?\nViral RNA extraction\n\nAdd ‘the’ before manufacturer’s.\n\nMicrocentrifuge ‘tubes’. Plural.\nHIV DRT\nAs this was a study validating an assay that relied heavily on reagent volumes, why aren’t the volumes of the original assay and their reaction temperatures not included here?\nReagent cost comparison\nAs the year is not out yet and conversion rates keep fluctuating, which conversion rate did you use?\nProvide references for definition of the following words: Precision, Reproducibility, Accuracy and amplification sensitivity\n\nStatistics\nSignificance was set at what p value?\nResults\nOnly two drug resistance genes are targeted in this assay, what informed the decision to arrive on these two sets of genes?\n\nTable 5 should be the one introducing the results subtopic. Introduce the samples, the viral loads before the validation assay results are presented.\n\nRemove the grids from the tables.\nDiscussion\n\nUse third person singular when discussing your results.\n\nYou don’t ‘show’ in discussion what has been shown in the results section.\n\nDelete ‘altogether’ in the first paragraph.\n\nWhat did Zhou et al.1 find out, and which procedure were they modifying? Include the year of the study too. As it has been used as the main reference, is it the only study that supports your findings? Are there studies that contradict your findings, and if there are, why aren’t they included here?\n\nIf the samples were obtained from patients who had not been exposed to protease inhibitors, why did your study focus on protease inhibitors?\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "53071", "date": "16 Oct 2019", "name": "Frank Onyambu", "expertise": [ "Reviewer Expertise HIV drug resistance and molecular genomics" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an innovative approach to reducing costs of offering HIV drug resistance testing in resource-limited settings. The authors have shown a thorough understanding of the literature in HIV drug resistance testing and implementation in resource-limited settings. They have related their findings to the majority of publications that presenting performance characteristics of in-house HIV drug resistance tests. It would be interesting for the authors to compare the performance of the ‘modified assay’ with the Viroseq assay as well.\n\nThe authors followed the World Health Organization (WHO) recommendations for validating 'in-house' HIV drug resistance assays and clearly showed that the results of the modified assay were comparable to the commercial assay. Furthermore, the authors have demonstrated that these assay modifications resulted in a significant decrease in cost of implementation. I however wonder why the authors did not present results for reducing reaction volumes for the RT-PCR step. It is conceivable that they may have at least attempted and therefore should comment if the results did not meet the WHO criterion for assay validation.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1518
https://f1000research.com/articles/8-1517/v1	28 Aug 19	{ "type": "Opinion Article", "title": "Repositories for academic products/outputs: Latin American and Chilean visions", "authors": [ "Leandro Torres", "Ricardo Hartley", "Leandro Torres" ], "abstract": "Open access policies have been progressing since the beginning of this century. Important global initiatives, both public and private, have set the tone for what we understand by open access. The emergence of tools and web platforms for open access (both legal and illegal) have placed the focus of the discussion on open access to knowledge, both for academics and for the general public, who finance such research through their taxes, particularly in Latin America. This historically unnoticed discussion must, we believe, be discussed publicly, given the characteristics of the Latin American scientific community, as well as its funding sources. This article includes an overview of what is meant by open access and describes the origins of the term, both in its philosophical sense and in its practical sense, expressed in the global declarations of Berlin and Bethesda. It also includes the notion of open access managed (or not) by some reputable institutions in Chile, such as CONICYT (National Commission for Scientific and Technological Research) and higher education institutions reputed nationally, such as the Universdad de Chile and Pontificia Universidad Católica de Chile. Various Latin American initiatives related to open access (Scielo, Redalyc, among others) are described, as well as the presence of Chilean documents in those platforms. The national institutional repositories are listed, as well as their current status and a discussion about what open access has implied in Latin America and its importance for the replicability of the investigations carried out locally. Finally, we describe some governmental initiatives (mainly legislative) at the Latin American level and propose some recommendations regarding the promotion and implementation of repositories for the access to scientific data (for access and replication purposes) of the national research.", "keywords": [ "Repositories", "Chile", "Latin America", "Scientific Data", "Open Access" ], "content": "Introduction\n\nInternationally, open access (OA) policies have gradually advanced in academia and beyond. Well known are national level OA policies and regulations from Peru1, Argentina2 and Mexico: Decreto Ley3, as well as global initiatives to remove obstacles for the access, distribution and re-use of academic research outputs in institutions that fund research, such as the National Institute of Health (NIH)4 and the National Science Foundation (NSF)5 in the United States, philanthropic organizations like the Bill and Melinda Foundation6, the European Commission7 and the Wellcome Trust in the UK8, among others (further information about OA at global level can be found in 9).\n\nAnother important element in the promotion of OA is the development of tools that can accelerate this process, that can direct to publications stored in OA repositories, such as Science Open, 1Science and web browser extensions such as Canary Haz10 and Unpaywall11, and the controversial academic social networks12 that allow self-archiving and distribution of publications such as academia.edu13 and researchgate14. This latter is currently under legal scrutiny for distributing academic publications15.\n\nIn the case of Latin America, while there has been important work in pushing some initiatives, some of which have shown progress, results do not seem very encouraging16,17. According to recent publication statistics, there is a linear increase in publications in OA journals; this is, however, rarely contrasted with the increase in the number of publications in non-OA journals in the region during 2013–2017, a trend that appears to be growing, at least according to indexes such as Web of Science. Therefore, the growth of OA has not had an effect on the number of publications in paywalled journals; instead, we see an increase in both and especially the latter.\n\nIn this analysis, it is necessary to take into account that OA refers only to documents that have been made available through ’Gold’ and ’Green’ OA modalities18. The former refers (although not in all cases) to OA articles for which publication is associated with article processing costs (APCs), while the latter makes reference to self-archiving, in which case a version of the peer-reviewed article is made available through an online repository or website, and archiving will depend on the policies that each journal imposes19.\n\nOA has been characterized by various authors, who have described its different varieties, among which we can identify: Libre OA20, Gratis OA20, Gold OA21,22, Green OA23, Hybrid OA24, Delayed OA:25 and Bronze OA26.\n\nIn addition, we should consider the following two non-traditional alternatives: Academic Social Networks (ASN), for-profit social networks that allow academics to share their publications, with more than half of their content being shared illegally; and Black OA, articles shared on illegal pirate sites such as SciHub. The data that Sci-hub has provided with respect to downloads from Chile (February 2016) show that these download concentrate mainly in the capital Santiago, where 273,834 articles were downloaded, followed by Concepción (31,985), Valdivia (22,069), Valparaíso (16,075) and Viña del Mar (17,024)27.\n\nAcademic Social Networks (ASN): corresponds to for-profit social networks that allow academics to share their publications. Although some include OA definitions12, others consider that the content shared through this platforms is not OA since, in contrast to Green OA repositories, these do not check copyrights and therefore, more than half of the its content is stored and shared illegally28, causing controversy29.\n\nBlack OA: corresponds to articles shared on illegal pirate sites, mainly Sci-Hub and LibGen29.\n\nRelated to Black OA, the data that Sci-hub has provided with respect to downloads from Chile (February 2016) show that these download concentrate mainly in the capital Santiago, where 273,834 articles were downloaded, followed by Concepción (31,985), Valdivia (22,069), Valparaíso (16,075) and Viña del Mar (17,024)27. Access to content through this kind of OA is common, as if all academic publications worldwide are considered, only 25% are disseminated through any existing form of OA, not including ASNs and black OA30.\n\nIn Latin America, historical OA movements such as SciELO, CLACSO, Redalyc and, more recently, LA Referencia, form the basis of what is understood by OA in the region. It has been argued that what “characterizes the Latin American flavour of OA” is that journals that constitute it are supported by universities, research institutes and other academic organizations without APCs31. However, obstacles in the interoperability between different search engines such as SciELO, Redalyc and LA Referencia31 hinder analyses of the necessary metrics that would allow a clear perspective on the advancement of the OA movement in the region. We must consider that, at least in Chile, the cost of collection subscriptions to publishers such as Elsevier, Springer-Nature, Wiley, American Chemical Society, Annual Reviews, Oxford University Press and AAAS from 2008 to 2017 was $95,754,011 USD for universities, government and other educational institutions, according to Consorcio para el Acceso a la Información Científica Electrónica,32, making it possible to reduce the cost of access to each article from $20 USD to $3.\n\nDespite these initiatives, out of the 20 countries considered part of Latin America, only Argentina, Perú and México have national laws that promote OA, especially through the development of institutional repositories (Perú1), Argentina2, México: Decreto Ley3), while others such as Colombia, Brazil and Chile have focused mainly on the management of national systems of digital repositories, despite the absence of mandatory policies for OA in national research.\n\n\nOA, a general vision\n\nThe philosophical roots of the information policies expressed as OA have their origins in the philosophy known as ’open society’, proposed by Henri Bergson (1859–1941), Karl Popper (1902–1994) and George Soros (1930), that propose fostering values such as freedom, progress, equality, fraternity, tolerance, rejection of tyranny, censorship and the exercise of power as a form of control.\n\nThe creation of institutions such as the Open Society Institute by George Soros (the name of which is inspired by the book ’Open Society and Its Enemies’ by Karl Popper, published in 1945), provided the context in which initiatives such as the Budapest Open Access Initiative33,34, the Berlin Declaration on Open Access to Knowledge in the Sciences and Humanities (2003)35 and Bethesda Statement on Open Access Publishing (also from 2003 and more oriented to research in the natural sciences)36 emerged.\n\nIn general, OA terms and (maybe even more importantly) access to research outputs, generally in the form of scientific publications, mean that social agents besides researchers are able to interact with such outputs, both in sciences and humanities. This democratization of knowledge creates a problem of access: it is not enough to make just the content available, it should also be interpretable by any of those who see it. It is maybe at this point that knowledge moves from being ’visible’ to being ’accessible’.\n\nAlthough this might seem obvious, it is of particular relevance if we consider that many of our individual decisions, as well as our collective political decisions, are many times based (or at least, should be based) on the knowledge that we have about certain phenomena. This point will be discussed later, when we reflect about what is understood by ’public good’, a fundamental element that often justifies the adoption of OA policies.\n\n\nThe vision of OA in Chile\n\nIn order to have a clear perspective of what OA entails at national level, we will take as reference the definitions of OA used by institutions associated with research in sciences and the humanities, taking as an example CONICYT (the largest funding body of science in Chile), Universidad de Chile and Pontificia Universidad Católica de Chile (both traditional universities and with high publication rates).\n\nThe first of these organizations is the main agency that funds and regulates academic research in Chile, and the other two are the most prestigious universities in terms of academic and research performance in the country, (highest QS ranked Chilean Universities in 2019).\n\nTo our knowledge, neither Conicyt nor Pontificia Universidad Católica de Chile have a clear definition of what is considered as OA; however, Universidad de Chile states on its website37:\n\n“The Open Access Movement is an initiative that promotes free access to digital materials derived from scientific or academic production, beyond copyrights, that these materials may hold.”\n\n“The idea is to enable people to read, download, copy, distribute, search or link these resources without needing to register or pay. This access is often performed through the internet and for this reason, the authors of these materials will receive a higher dissemination of their work. These digital contents can include articles published in online journals, images, data, audiovisual material, and any other digital content whose author wishes to give free access”.\n\nTo comply with these objectives, they propose options to publish in OA journals, among which they include the 5000 magazines present in the Directory of Open Access Journals (DOAJ), and institutional repositories such as Universidad de Chile’s repository CAPTURA, CONICYT’s repository and SHERPA/RoMEo38.\n\nWillisky39 and Fischman40 propose 10 different ’flavours’ of OA that correspond to different strategies oriented towards the diffusion and promotion of academic knowledge. Those correspond to 10 different “flavours” of Open Access are: Home page (websites in which a profile of the academics and their publications are made available), E-print (academic repository that allows self archiving published and unpublished material), Author fee (fees support to immediate and complete access to journals or individual paid articles), Subsidized (different institutions enable complete access to open access journals), dual mode (there is a subscription to the print edition that is used to support the digital and printed formats), Delayed (subscription fees are collected for immediate access and to sustain a printed edition, the content is fully available after a period of time), Partial (a part of the content is open access whereas the rest require a subscription). These flavours are specially used by publishers from developed countries to promote OA. In contrast, developing countries as Chile usually adopt Per capita subscriptions (expense limited to registering institutions in an access management system, such as Cincel and others) and Indexing (access to abstract and bibliographic information provided as a government service, with access to whole articles obtained by paying per view). Finally, a certain amount of the academic production in Chile is supported by academic institutions affiliated to SciELO Chile. This last modality is known as ’Cooperative’.\n\n\nLatin American OA initiatives and making Chilean research visible\n\nLatindex (Latin American Index of Serial Scientific Publications): a system for online dissemination that gathers information about scientific research journals, professional magazines and magazines dedicated to scientific and cultural outreach, from countries in Latin America, the Caribbean, Spain and Portugal. It emerged as an initiative from the National Autonomous University of Mexico (UNAM) in 1995, becoming a network for regional cooperation since 1997. It offers data related to print or online journals that comply with academic criteria (in a broader sense) of quality that Latindex has established, as well as access to full academic journals and articles in the different languages used in Latin America. It is the most inclusive source41, since from its origins it used files present in the platforms CLASE, PERIODICA and LILACS, although currently it has a broader coverage. For Latindex, the identification, registry and update of its entries has been difficult due to the different criteria and standards for publication. However, it has been able to gather fundamental data about publications in the affiliated regions. In 2018, Latindex registered 26,010 journals, of which 15,948 are Latin American (South America, Central America and the Caribbean). Of those, 2,147 are Chilean, which represents 8,251% of the total; among these, 2,038 are still being issued, 105 are not being published anymore and four have an unknown status41.\n\nLatindex provides users access to a directory that takes into account traditional journals with an international distribution as well as newer journals with limited distribution. The criteria to be included in the directory are the following (https://www.latindex.org/latindex/regRev): The journal must be at least two years old, item The official or institutional website of the journal will be assessed and must have free access to all its contents, non one pdf journal, among others.\n\nSciELO (Scientific Electronic Library Online): a scientific library with a digital database of academic articles belonging to different disciplines, freely available in full text, which allows for preparation, outreach, storage and evaluation of academic literature in electronic format. It works under a set of common criteria of publication and software, and its operation is based on national collections of academic journals with editorial committees and is peer reviewed. One of its main purposes is to make visible academic articles in local languages (Meneghini 2006). By 2018, it indexed 1,285 journals, 145,182 academic articles and 16,943,454 citations (confirma SciELO en números). This initiative was developed by São Paulo Research Foundation (Fundação de Amparo à Pesquisa do Estado de São Paulo FAFESP) and the Latin American and Caribbean Center on Health Sciences Information (BIREME). SciELO Chile indexes 121 journals, of which 107 are still being issued and 14 are no longer published. Each SciELO website is responsible for the costs and responsibilities of obtaining citation data of the indexed articles. In some cases, this is performed by the academic journals and in others, the organization in charge of SciELO42.\n\nRedalyc: a more recently created indexing and publication platform. This network publishes more than 1,278 academic journals from Ibero America and the Caribbean. It stores about 47,056 individual issues and 609,283 full text articles and documents of diverse nature, being the database that provides the largest amount of metadata associated to authorship and co-authorship in Latin America. With respect to Chile, RedALyC indexes 93 journals, 37,921 full articles and 110 instalments. The catalogue of RedALyC selects journals in the social sciences and humanities based on a set of criteria (for details see 43). Despite its high level of access to academic documentation, RedALyC still lacks a citation processing system and impact indications in the region. Its broad access to journals in the region has allowed it to harvest a large amount of metadata associated with publications, which represents an important potential source for analysis in issues related to academic collaboration both inside and outside the region, using what they refer as “production profiles”44.\n\nCLACSO (Latin American Council of Social Sciences): a non-governmental international organization created in 1967 in which 47 countries participate (including United States, Canada, Germany, France, Portugal and Latin American countries) with a total of 616 research centres in the social sciences and the humanities. It has generated initiatives that promote the development of OA in different regions through the publication of books, journals and other formats through its virtual library45, the Latin American and the Caribbean Library of Social Sciences and CLACSO TV46.\n\nLA Referencia: The Federated Network of Institutional Repositories of Scientific Publications, LA Referencia, was created on the 29 of November, 2012, after the signature of a cooperation agreement in Buenos Aires47, with the purpose of providing OA to academic research financed through public funds in the region. It is a network of OA repositories present in nine countries (Argentina, Brazil, Chile, Colombia, Ecuador, El Salvador, Mexico, Peru and Costa Rica). This platform aims to generate an interoperable system, in which it would be possible to share and make visible academic outputs generated in higher education institutions and scientific research organizations. It is operated by national nodes, storing and hosting academic articles and post-graduate theses. It is based on agreements between public agencies of science and technologies (ministries and national agencies of science and technology) of the member countries, together with red-CLARA (Latin American Cooperation of Advanced Networks), whose objective is to connect academic computational networks in Latin America48,49. Its purposes include (1) to realize regional agreements to define common standards and policies, (2) to ’harvest’ the registries of metadata obtained from each national node, (3) to validate the quality of the data obtained, (4) to generate a unified search service (federate) and (5) to facilitate access to full text documents and bibliometric data50. The national node corresponding to Chile has a council of directors belonging to the National Commission of Scientific and Technological Research (CONICYT), and its function is “to strengthen and ensure the access to national and international scientific information with the purpose of research, education and innovation”, by managing the national infrastructure of access to STI information. The policy and regulation that supports its actions is the ’proposal for open data’ generated by the government of Chile in 2014, which introduces seven relevant aspects that recipients of CONICYT grants should consider, such as: publish data (non mentioning raw data) and other products in institutional repositories, provide the location to CONICYT, publish before one year from final report, CONICYT “should” publish final report not longer than 3 months after the approval.\n\nIn this respect, CONICYT provides recommendations for OA and preserving scientific information and data in the ’Manual for open data’51. As a platform, it is managed by Web SIC (Scientific Information System) that hosts CONICYT’s Digital Repository, in which it is possible to store and access the results of research, productivity and instruments financed by this agency. The Chilean institutions that participate in LA Referencia can be found on SciELO-Chile and their Institutional Repository.\n\nStandardization, regional and international interoperability, regional collection of data and training are some of the main objectives of this initiative, that work together with The Confederation of Open Access Repositories (COAR)52.\n\n\nChilean initiatives: OA repositories and journals\n\nThe Chilean academic heritage is published and preserved through national efforts of different kinds of institutions, such as universities and government organizations. The efforts have been directed towards the construction of institutional repositories (IRs) and journals that feature OA licenses (OAJ). These store, share and provide access to academic products from institutions and their researchers. This research can be made available in various formats, from theses, monographs and preprints to published documents (in OA journals or with distribution licenses).\n\nThree important initiatives have emerged as a way of organizing and making visible global OA content: Directory of Open Access Repositories (OpenDOAR, lists OA repositories), Directory of Open Access Journals (DOAJ, where OA journals are indexed) and Registry of Open Access Repositories (ROAR, indexes the growth of institutional OA repositories and OA policies).\n\nIn ROAR9, Chile registers 22 entries, of which 16 belong to higher education institutions, one is SciELO Chile, one the Digital Library of the Museum of Memory and Human Rights and one the digital repository of CONICYT. The rest of the repositories correspond to duplicated or inactive entries.\n\nOf the 22 repositories, 13 correspond to institutional repositories of Chilean universities, four correspond to institutional repositories of government bodies and 13 operate using the DSpace software for the creation and management of digital repositories.\n\nIn DOAJ, Chile registers 113 OA journals, most of them corresponding to publications related to the social sciences. Among these, 109 do not charge APCs. With respect to licenses, 39 use the licence CC-BY, 26 the licence CC BY-NC-ND, 17 use the license CC BY-NC-SA, 15 the licence CC BY-NC, 12 the licence CC BY-SA (12), and only three use specific editorial licenses (further details about the editorial licenses can be found in 54).\n\nDespite the high number of institutional repositories, it seems as if there is no explicit recognition from the authorities and national organizations regarding academic research in relation to this matter, since there are no mandated policies with respect to the need, construction and implementation of this type of repository.\n\n\nThe implications of worldwide OA and how has it been understood in Latin America: the importance of replicability and reproducibility of science\n\nAs mentioned above, and according to the Budapest Declaration, OA is defined as an editorial model where access and use of scientific literature is free55,56). In Latin America, there are various interpretations of the meaning of OA and the consequences of accepting some of these meanings, due to different motivations for participation in this movement57,58. Whatever its meaning for the organizations in charge of this initiative (national science and technology agencies, universities, editorials, etc.) in Latin America, it is clear that currently, two ways of OA are recognized: ’Green OA’, in which the papers that have been published under a traditional model or that are in the process of being published are made available in repositories; and ’Gold OA’, in which articles are published in journals that have ’opening’ mechanisms in their publications, generally with publications costs for the researcher or its founders. In either case, the method of access is through the internet.\n\nThe impact of both methods has been recently tested and it was been concluded that only 13% of the articles published between 2008 and 2015 are available through Green OA, while 9% were available through Gold OA, showing that most of what is being published is being hidden under a ’paywall’59, making this knowledge practically inaccessible to researchers and institutions that do not have large funding sources and even more difficult in the case of access for educational purposes and the general public.\n\nWhile so far we have stressed the importance of OA for citizens, an important aspect, and probably essential for science itself (and the humanities)60–64 is the possibility to critically analyze the corrections of scientific declarations and the conclusions that other scientists have reached and their data analysis, both at systematic and statistical levels65). To fulfill this task, it is necessary that other researchers can reuse, replicate and reproduce the findings reported in scientific publications. This terminology has been a source of confusion (to learn more about this confusion, see 66). Goodman67) proposes the following definitions of reproducibility:\n\nMethods reproducibility: provide sufficient detail about procedures and data so that the same procedures could be exactly repeated.\n\nResults reproducibility: obtain the same results from an independent study with procedures as closely matched to the original study as possible.\n\nInferential reproducibility: draw the same conclusions from either an independent replication of a study or a reanalysis of the original study.\n\nTo achieve reproducibility, it is necessary that researchers share materials, protocols and data that sustain findings, with the purpose of confirming previous studies or becoming the basis for new studies. For example, Nature Cell Biology68 proposes that the minimum combination of data would include not only the data presented in the publication itself, but also the unprocessed numerical data that underlie the graphs and quantitative evaluations, the independent repetitions of representative experiments that provide the support for reproducibility in the results and sets of large scale data generated by the study69.\n\nThese data should be available with the published document (including the original data without processing and independent repetitions). However, larger datasets could be stored in public databases and their location specified in the corresponding document. In the absence of specific OA databases (for example, for new techniques or for which there is still no public database), Nature recommends to store data in general data repositories, such as Figshare70 or Dryad71. This databases can also be complemented, depending on its appropriateness, with tools like Zenodo72, Github73, Gitlab74 or OpenScience Framework75, among others, as platforms for publication, as well as for preprints, data and supplementary code resulting from a research project.\n\nDespite the relevance of OA databases, a study performed in 318 biomedical research76 journals found that, as reflected in the instructions for authors and their editorial policies, only a small percentage (11.9%) declare explicitly that sharing data is a necessary condition for publication, 9.1% requested shared data (without this being an explicit condition for publication), 23.3% of journals promote sharing data but not as mandatory, 9.1% indirectly mention the relevance of sharing data and 14.8% propose to share data of proteins, proteomics and genomics. The 31% left does not mention data sharing practices. Although, as the authors mention, 65,7% of the journals require data sharing as a criteria for replicability, they do not provide specific policies or guidelines about these practices that could ensure that the data is available and reusable.\n\nIn Latin America, and in particular in Chile, little emphasis has been given to the importance of access to data or the development and promotion of platforms to store this type of research output with the purpose of promoting reproducibility. A Latin American study analyzed whether there are differences in the beliefs that regulate research between natural and social scientists77. The article examines the social agreements that regulate the behaviour of academic researchers through an opinion survey in which 185 active researchers participated (of which, 96% declared to have completed postgraduate studies).\n\nThis study proposes four fundamental (intuitive and informal) agreements that constitute scientific work, including:\n\nThe world has laws or regularities that are understood through observation\n\nYou must have the ability to analyze in an objective, impartial, verifiable and systematic way the information provided by reality (\"having a critical attitude\").\n\nHave mastery of technical aspects of the work, the correct use of the recording devices, the baseline calibration for each experiment, the design of appropriate controls (\"have methodological aptitude\"), among others.\n\nThe results should be communicated in an open way, that is, verifiable or replicable.\n\nThe most remarkable element of this work in terms of OA (as a means of communication of science) is precisely that expressed in the last point, since the interviewees (96% natural scientists 91 social scientists) mention that the importance of ensuring the veracity and verifiability of the data reported by an investigation are necessary for the advancement of knowledge.\n\nAlthough this research does not discuss how to achieve reproducibility through the use of OA databases, it is one of the few precedents in this topic in Latin America. Additionally, in some blogs on the SciELO website, the problem of replicability has been superficially discussed. In these blog entries, Lilian Nassi-Caló78,79 discusses what has been going on with the so-called ’reproducibility crisis’ of certain scientific disciplines, but makes no mention to what this implies for the future of Latin American science. The only claim that CONICYT has made on this topic was to define what is understood as scientific knowledge, including in this definition the idea of reproducibility (without a deeper discussion of the concept):\n\n“Scientific knowledge, generated in this way, has the qualities of reproducibility and objectivity. Precisely herein resides an essential part of its enormous utility, since its predictive value applies to all situations in which the established conditions are reproduced, despite the subjectivity of the observer”80.\n\nAs mentioned, Chile does not have clear perspectives or statements about public databases and their relevance, not only for OA itself, but also for reproducibility of Chilean and international science, since it is possible that in the near future, what we today call impact factor will be highly conditioned by the reproducibility of the results of research.\n\nIn this light, it seems evident that the exchange of data and its accessibility is fundamental to the advancement of knowledge in the natural sciences, social sciences and humanities, since it allows validation and to contribute with new knowledge about information already published. This supports the development of research and innovation at national and international levels, together with democratization of access to such data or registries in many cases. All this will happen only if this information is available, in an organized and coherent way, according to the declarations of data availability. This can be observed both in prescriptions of governmental mandates, as well as editorials in charge of access and dissemination of the outputs of research. To learn more about this, please see 68 and 81–83.\n\n\nReflections about the current state of OA in Chile\n\nAs mentioned above, OA and the availability of resources are important, not only for the development of new knowledge, but also for social, economic and political advancement in a democratic society. If we only analyze South American countries, according to ROARMAP, only 49 policies have been adopted in this region, of which seven correspond to Argentina, one to Bolivia, six to Colombia, eight to Perú and four to Venezuela. Brazil leads the list with 23 policies84. Most of these policies have been developed by research organizations. Despite Latin American efforts, to date (March 2019), Chile is not present in this database.\n\nAs we discussed in this paper, although there are some Latin American initiatives that support OA, mainly in the use of licenses and the development of international databases (Scielo, Redalyc, CLACSO library, LA Referencia), it is important to note that there are still important challenges for interoperability and access to documents (access in terms of a platform or a system enabling the quick and precise location of a document, and interoperability in terms of a systematization of metadata from these different initiatives).\n\nIt is important to highlight that university repositories are an essential element in the diffusion and communication of knowledge. The number of digital documents that universities systematically produce, considering academics, undergraduate and postgraduate students, are elements that can help us define the profile of the university, regional, national and international research, and its opportunities for collaboration. This goes far beyond simple university outreach and public engagement activities, and should be treated as a research process in itself. This is why it is an essential element of the academic cycle, implying a responsibility for the administration in the development of technological capacities, as well as human resources within the institution (here, we call it human resources since the type of collaboration required for this purpose is interdisciplinary and does not limit itself to academic titles or degrees, an aspect poorly developed in Chile). This is not a new revelation; in fact, it has been mentioned by the Association of Research Libraries85, among others.\n\nA fundamental role of universities is the dissemination of knowledge and the recognition of intellectual capital and knowledge produced in them. For this, they must develop strategies that guarantee the distribution of the generated content, which implies that they must recover the ability to manage their intellectual capital in order to promote the resolution of local (or global) problems. Therefore, it is important that local governments promote (and finance) these types of initiatives.\n\nThe latter is of particular relevance considering the current state of OA worldwide and the recent emergence of Plan S, an initiative developed by cOAlition in 2018 and revised in 2019 and that proposes 10 principles so that in 2021, publicly funded research is published exclusively in OA journals. Several of the principles are ambiguous and inaccurate and some controversial, such as the impediment of researchers to publish in hybrid journals, preventing researchers from publishing in about 80% of academic journals, including Nature, Science, The Lancet, etc. In response to this, more than 1500 researchers, including two Nobel Prize winners, signed a letter arguing against the plan. Agostini and Berk85 point out various points of relevance when considering the current status of Plan S, important elements to be considered by Latin American countries for approval by researchers in Latin American financing agencies, including: (1) quality versus number of publications (the OA business model prioritizes quantity over quality, favoring the appearance of predatory journals, which generates waves of misinformation); (2) increases in the costs associated with publication that could only be financed by governments and richer institutions, generating a gap of inequality; (3) segregation of research quality, diversity and isolationism (countries that generate the highest amount of high quality publications will continue to publish in high-impact journals while researchers from emerging countries will not); (4) would compromise the peer review process (high publication costs would reduce the number of researchers willing to review for “inequality aversion” free of charge, but if the reviewers are paid it would increase the cost for researchers who publish in open journals); and (5) a reproducibility crisis (total open access will facilitate publication with lower quality thresholds, the public would be exposed to a world of information that is more accessible but less accurate and easier to manipulate). In particular for this last point, we point out that the use of models and knowledge management platforms that combine being reader-friendly with a depth of detail that allows critical reading by specialists and non-specialists is required. It is evident that the countries of Latin America will be at a disadvantage in the face of these type of measures, considering that most of the research funding is through public funds. This is particularly important if we think in the little funding (or lack in some cases) to promote open access in South America and the redundancy of initiatives trying to ’harvest’ publications and their metadata and trying to make different platforms ’interoperable’. We propose instead that the development and maintenance of institutional repositories and the use of already existing platforms should be prioritized.\n\nOur proposal aims to promote and strengthen the use of institutional repositories as disciplinary (national and/or global), in parallel to the efforts made to increase the quality of research in the region mainly in strengthening the reproducibility and/or replicability of the investigations developed.\n\nWe should also consider what is indicated in the document OpenUP – 7107220 Deliverable D4.3 Good practices and lessons learned. Briefly, research dissemination, as defined by Wilson [? ], facilitates research uptake and understanding. Furthermore, when implementing an institutional repository in Latin America and, in particular, in the Chilean context, it is necessary to explore and focus efforts on the prescriptions provided by institutions with a well-defined trajectory, such as The Repositories Support Project (RSP), an initiative created to “contribute to the creation of capacities, knowledge and abilities in higher education institutions of the United Kingdom”84. RSP has developed a repertoire of considerations to select the best platform to be implemented, taking into account technical aspects86, metadata, repository management, outreach and user engagement and a checklist87.\n\nAlso important, on a prescriptional basis, is the use of ’quality seals’ with respect to repositories, such as CoreTrustSeal88, based on DSA-WDS Core Trustworthy Data Repositories Requirements89. These initiatives seek to generate an adequate provision of (1) services for authors and editors (2) deposit, treatment and long term storage of documents and metadata of the objects stored (3) public availability of the objects, which guarantees the possibility of access for humans and machines (necessary for complementary and integral services) and (4) the transference of metadata.\n\nLastly, we propose a ’Guide for the evaluation of institutional research repositories’ that, together with the directives of OpenAIRE90, considers the following aspects:\n\nVisibility\n\nPolicies\n\nLegal aspects\n\nDescriptive metadata for publication (OAI-DC)\n\nLogs and statistics\n\nSecurity, authenticity and integrity of data\n\nServices and added value functionality\n\nWe should also take into account aspects associated with national and institutional intellectual property policies. Usually, with respect to scientific articles, most scientific journals request the transfer of copyrights to the journal, whose consequences includes the loss of rights to publish of one owns work on a personal website without permission of the editor and the inability to provide copies of one’s own work for distribution and utilization as an educational tool or in the development of academic curricula91. To remedy these problems, an important measure is to modify the agreements provided by journal editors. For this reason, SPARC (Scholarly Publishing and Academic Resources Coalition) has created an Appendix for authors that can be attached to the publication agreement of journals92.\n\nWe believe that any institutional repository should allow access, not only to the document in a legible format such as a PDF of the print version, but also should allow access for technologies such as data mining and other free services, which would promote reuse and add value to research already conducted. This could be achieved through the use of repositories that fulfill the minimum requirements to be incorporated in services such as CORE93, which collect all OA content from different sources for its analysis. In this sense, and as a way to generate added value and true access for all citizens that finance, through public funds, these initiatives49, we propose to conduct an exhaustive bibliometric analysis that will allow researchers to establish narratives of conservation, territorial and urban planning, and more. This is not new, but is one of the outreach strategies proposed by Rogers (1962). According to Wilson94, diffusion of innovation “offers a theory of how, why, and at what rate practices or innovations spread through defined populations and social systems\". Also OpenUP – 7107220 Deliverable D4.3 Good practices and lessons learned95 should be considered a central element, since it shows that the initiatives mentioned earlier are not just mirrors of academic activities that ’make it visible’, but it also enable access from different social actors towards local knowledge (collaborative initiatives of citizen science, interaction between industry and academics, etc.), and it allows more rapid advancement in addressing local and Latin American needs in general. To achieve this, it is urgent that Chile follows the countries (Mexico, Argentina and Peru) that have established mandatory OA policies for their own national academic research.\n\n\nData availability\n\nNo data are associated with this article.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nhttps://portal.concytec.gob.pe/images/stories/images2013/portal/areas-institucion/dsic/ley-30035.pdf. [date retrieved 05/10/19].\n\nhttps://www.boletinoficial.gob.ar/#!DetalleNormaBusquedaAvanzada/98996/20131209. [date retrieved 05/10/19].\n\nhttp://www.dof.gob.mx/nota_detalle.php?codigo=5345503&fecha=20/05/2014. [date retrieved 05/10/19].\n\nhttps://publicaccess.nih.gov/index.htm. [date retrieved 05/10/19].\n\nhttp://ec.europa.eu/research/participants/data/ref/h2020/grants_manual/hi/oa_pilot/ h2020-hi-oa-pilot-guide_en.pdf. [date retrieved 05/10/19].\n\nhttps://www.nsf.gov/pubs/2015/nsf15052/nsf15052.pdf. [date retrieved 05/10/19].\n\nhttp://www.gatesfoundation.org/How-We-Work/General-Information/Open-Access-Policy. [date retrieved 05/10/19].\n\nhttps://wellcome.ac.uk/press-release/wellcome-trust-strengthens-its-open-políticadeacceso. [date retrieved 05/10/19].\n\nroarmap.eprints.org. [date retrieved 05/10/19].\n\nhttps://labworm.com/tool/canary-haz. [date retrieved 05/10/19].\n\nhttps://unpaywall.org/. [date retrieved 05/10/19].\n\nBjörk BC: The open access movement at a crossroad: are the big publishers and academic social media taking over? Learn Publ. 2016b; 29(2): 131–134. Publisher Full Text\n\nhttps://www.academia.edu/. [date retrieved 05/10/19].\n\nhttps://www.researchgate.net/. [date retrieved 05/10/19].\n\nhttp://www.responsiblesharing.org/coalition-for-responsible-sharing-status-report-on-researchgate/. [date retrieved 05/10/19].\n\nhttp://biblioteca.clacso.edu.ar/clacso/biblioteca/20170719025829/Open_Access_in_Latin_America.pdf. [date retrieved 05/10/19].\n\nhttp://biblioteca.clacso.edu.ar/clacso/se/20150722110704/HechoEnLatinoamerica.pdf. [date retrieved 05/10/19].\n\nHarnad S, Brody T, Vallières F, et al.: The Access/Impact Problem and the Green and Gold Roads to Open Access: An Update. Serials Rev. 2008; 34(1): 36–40. Publisher Full Text\n\nSherpa.ac.uk.romeo. [date retrieved 05/10/19].\n\nSuber P: Gratis and libre open access. 2008; [date retrieved 05/10/19]. Reference Source\n\nArchambault E, Côté G, Struck B, et al.: Research impact of paywalled versus open access papers. 2016. [date retrieved 05/10/19]. Reference Source\n\nGargouri Y, Larivière V, Gingras Y, et al.: Green and gold open access percentages and growth, by discipline. 2012. Reference Source\n\nHarnad S, Brody T, Vallières F, et al.: The Access/Impact Problem and the Green and Gold Roads to Open Access: An Update. Serials Rev. 2008; 34(1): 36–40. Publisher Full Text\n\nLaakso M, Björk BC: Delayed open access: an overlooked high-impact category of openly available scientific literature. J Am Soc Inf Sci Tec. 2013; 64(7): 1323–1329. Publisher Full Text\n\nWillinsky J: The access principle: the case for open access to research and scholarship. 2009. Reference Source\n\nPiwowar H, Priem J, Larivière V, et al.: The state of OA: a large-scale analysis of the prevalence and impact of Open Access articles. PeerJ. 2018; 6: e4375. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nhttp://www.sciencemag.org/news/2016/04/whos-downloading-pirated-papers-everyone. [date retrieved 05/10/19].\n\nJamali HR: Copyright compliance and infringement in ResearchGate full-text journal articles. Scientometrics. 2017; 112(1): 241–254. Publisher Full Text\n\nChawla D: Publishers take researchgate to court, alleging massive copyright infringement. 2017; [date retrieved 05/10/19]. Reference Source\n\nTennant JP, Waldner F, Jacques DC, et al.: The academic, economic and societal impacts of Open Access: an evidence-based review [version 3; peer review: 4 approved, 1 approved with reservations]. F1000Res. 2016; 5: 632. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBabini D, Machin-Mastromatteo JD: Latin American science is meant to be open access: Initiatives and current challenges. Information Development. 2015; 31(5): 477–481. Publisher Full Text\n\nAssociation of Research Libraries: Consorcio para el acceso a la información científica electrónica. 2018. [date retrieved 05/10/19]. Reference Source\n\nBrenda E, Chavez A: Leyes de acceso abierto: estudio comparado entre mexico, argentina y peru. Tesis de maestría en bibliotecología y estudios de la información. Universidad nacional autónoma de méxico. posgrado en bibliotecología y estudios de la información. 2017. [date retrieved 05/10/19]. Reference Source\n\nhttps://www.budapestopenaccessinitiative.org/boai15-1. [date retrieved 05/10/19].\n\nhttps://openaccess.mpg.de/Berlin-Declaration. [date retrieved 05/10/19].\n\nhttp://legacy.earlham.edu/~peters/fos/bethesda.htm. [date retrieved 05/10/19].\n\nOpen access. [date retrieved 05/10/19]. Reference Source\n\nTipos de publicaciones. [date retrieved 05/10/19]. Reference Source\n\nWillinsky J: The access principle: The case for open access to research and scholarship. Cambridge, ma: Mit press, 2006. Reference Source\n\nAlperin JP, Fischman GE, Willinsky J: Open access and scholarly publishing in latin america: ten flavours and a few reflections. Liinc em Revista. 2008; 4(2): 172–185. Reference Source\n\nhttp://latindex.org/latindex/tablaPais?id=13&id2=0. [date retrieved 05/10/19].\n\nhttps://scielo.conicyt.cl/scielo.php?lng=es&script=sci_alphabetic. [date retrieved 05/10/19].\n\nhttp://www.redalyc.org/redalyc/media/redalyc_n/estaticasredalyc/Criterios/criterios.html. [date retrieved 05/10/19].\n\nhttp://ri.uaemex.mx/handle/20.500.11799/573. [date retrieved 05/10/19].\n\nhttp://biblioteca.clacso.edu.ar/. [date retrieved 05/10/19].\n\nhttps://www.clacso.org.ar/acceso_abierto_ y_difusion/presentacion.php?s=10&idioma. [date retrieved 05/10/19].\n\nhttp://www.lareferencia.info/es/recursos/documentos/acuerdos-politicos/2-acuerdo-de-cooperacion-regional-acta-de-buenos-aires-que-constituye-la-referencia-2012. [date retrieved 05/10/19].\n\nhttps://www.redclara.net/index.php/es/. [date retrieved 05/10/19].\n\nhttp://www.lareferencia.info/es/. [date retrieved 05/10/19].\n\nhttp://www.lareferencia.info/es/institucional/el-buscador. [date retrieved 05/10/19].\n\nhttp://datoscientificos.cl/politicabhttp://datoscientificos.cl/politica/documentos. [date retrieved 05/10/19].\n\nhttp://www.unesco.org/new/en/communication-and-information/portals-and-platforms/goap/key-organizations/latin-america-and-the-caribbean/coar/. [date retrieved 05/10/19].\n\nhttp://v2.sherpa.ac.uk/view/repository_by_country/cl.html. [date retrieved 05/10/19].\n\nhttps://creativecommons.org/licenses/?lang=es. [date retrieved 05/10/19].\n\nCartes-Velásquez R: Open access in chilean dentistry journals. Int J Odontostoma. 2010; 4(2): 123–126. [date retrieved 05/10/19]. Publisher Full Text\n\nSánchez-Martín FM, Millán Rodríguez F, Villavicencio Mavrich H: La iniciativa open access (oai) en la literatura científica. Actas Urológicas Españolas. 2009; 33(7): 732–740. [date retrieved 05/10/19]. Reference Source\n\nVillanueva E: Accidental open access and the hazards involved: Preliminary experiences on internet-based publishing in a peruvian university. First Monday. 2006; 11(6). Publisher Full Text\n\nTerra Figari LI: Diseminación del conocimiento académico en américa latina. montevideo. in antropología social y cultural en uruguay 2007. Uruguay: Unesco, 2008. Reference Source\n\nZhang L, Watson EM: Measuring the impact of gold and green open access. J Acad Libr. 2017; 43(4): 337–345. [date retrieved 05/10/19]. Publisher Full Text\n\nPeels R, Bouter L: The possibility and desirability of replication in the humanities. Palgrave Commun. 2018; 4: 95. [date retrieved 05/10/19]. Publisher Full Text\n\nCamerer CF, Dreber A, Holzmeister F, et al.: Evaluating the replicability of social science experiments in nature and science between 2010 and 2015. Nat Hum Behav. 2018; 2: 637–644. [date retrieved 05/10/19]. Publisher Full Text\n\nPeels R, Bouter L: Humanities need a replication drive too. Nature. 2018; 558(7701): 372. [date retrieved 05/10/19]. PubMed Abstract \| Publisher Full Text\n\nde Rijcke S, Penders B: Resist calls for replicability in the humanities. Nature. 2018; 560(7716): 29. [date retrieved 05/10/19]. PubMed Abstract \| Publisher Full Text\n\nPeels R, Bouter L: The possibility and desirability of replication in the humanities. Palgrave Commun. 2018; 4: 95. [date retrieved 05/10/19]. Publisher Full Text\n\nPlesser HE: Reproducibility vs. Replicability: A Brief History of a Confused Terminology. Front Neuroinform. 2017; 11: 76. [date retrieved 05/10/19]. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDrummond C: Replicability is not reproducibility: Nor is it good science. 2009. [date retrieved 05/10/19]. Reference Source\n\nGoodman SN, Fanelli D, Ioannidis JP: What does research reproducibility mean? Sci Transl Med. 2016; 8(341): 341ps12. [date retrieved 05/10/19]. PubMed Abstract \| Publisher Full Text\n\nhttps://www.nature.com/sdata/policies/repositories. [date retrieved 05/10/19].\n\nOn data availability, reproducibility and reuse. Nat Cell Biol. 2017; 9(4): 259. [date retrieved 05/10/19]. PubMed Abstract \| Publisher Full Text\n\nhttps://figshare.com/. [date retrieved 05/10/19].\n\nhttps://dryad.figshare.com/. [date retrieved 05/10/19].\n\nhttps://zenodo.org/. [date retrieved 05/10/19].\n\nhttps://github.com/. [date retrieved 05/10/19].\n\nhttps://about.gitlab.com/. [date retrieved 05/10/19].\n\nhttps://osf.io/. [date retrieved 05/10/19].\n\nVasilevsky NA, Minnier J, Haendel MA, et al.: Reproducible and reusable research: are journal data sharing policies meeting the mark? PeerJ. 2017; 5: e3208. [date retrieved 05/10/19]. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nde la Lama García A, del Castillo Mussot M, de la Lama Zubirán MA: ¿Existen diferencias en las creencias que regulan las investigaciones de los científicos naturales y sociales? 185 investigadores responden. Argumentos (Méx.) 2013; 26(71). [date retrieved 05/10/19]. Reference Source\n\nNassi-Caló L: La reproducibilidad en los resultados de investigación: la punta del iceberg. SciELO en Perspectiva. 2014; [date retrieved 05/10/19]. Reference Source\n\nNassi-Caló L: La reproducibilidad en los resultados de investigación: la mirada subjetiva. SciELO en Perspectiva. 2014; [date retrieved 05/10/19]. Reference Source\n\nCONICYT: Ciencia y tecnologÍa en chile: ¿para quÉ? 2010; [date retrieved 05/10/19]. Reference Source\n\nhttps://www.nature.com/authors/policies/ availability.html#data. [date retrieved 05/10/19].\n\nhttp://www.loadb.org/Control.do?_brse. [date retrieved 05/10/19].\n\nhttps://www.nature.com/articles/s41562-016-0021. [date retrieved 05/10/19].\n\nhttp://roarmap.eprints.org/view/country/019.html. [date retrieved 05/10/19].\n\nAgustini B, Berk M: The open access mandate: Be careful what you wish for. Aust N Z J Psychiatry. 2019; 4867419864436. PubMed Abstract \| Publisher Full Text\n\nRSP United Kingdom: About repositories support project. [date retrieved 05/10/19]. Reference Source\n\nRSP United Kingdom: Technical approaches. [date retrieved 05/10/19]. Reference Source\n\nCoretrustseal certification. [date retrieved 05/10/19]. Reference Source\n\nDsa-wds core trustworthy data repositories requirements. [date retrieved 05/10/19]. Reference Source\n\nOpenaire guidelines for literature repositories. planning checklist. [date retrieved 05/10/19]. Reference Source\n\nUc berkeley, faculty conference on scholarly publishing. taking back control: Managing copyright and intellectual property. [date retrieved 05/10/19]. Reference Source\n\nSPARC: Author addendum to publication agreement. [date retrieved 05/10/19]. Reference Source\n\nCORE: Aggregate all open access research outputs from repositories and journals worldwide and make them available to the public. [date retrieved 05/10/19]. Reference Source\n\nWilson PM, Petticrew M, Calnan MW, et al.: Disseminating research findings: what should researchers do? A systematic scoping review of conceptual frameworks. Implement Sci. 2010; 5: 91. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nhttp://openup-h2020.eu/wp-content/uploads/2018/10/OpenUP_D4.3_Innovative_dissemination_methods_Good_practices_and_lessons_learned.pdf. [date retrieved 05/10/19]." }	[ { "id": "53091", "date": "17 Sep 2019", "name": "Claudio Gutierrez", "expertise": [ "Reviewer Expertise data" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article is mostly a survey (with a discussion) of the state of Open Access Policies in Latin America and in Chile.\nThis is a relevant topic on which little systematic literature exists. Thus it is an important scientific contribution that deserves to be published.\nThe material of the article is interesting, the authors have done a lot of work compiling sources and developing a map of the realm of Open Access in Latin America and in Chile.\nI only have formal observations, mainly regarding the organization of the paper that I will list below.\nA. I suggest to organize the material according to the three main topics treated (hopefully in this order) that currently are spread and mixed in different parts of the paper:\nOpen Access Policies and notions (varieties of OA, reproducibililty, legal issues, etc.). (page 3, page 7, page 9)\n\nOA in Latin America (end of page 4, page 5, part of page 7, parts of page 9)\n\nOA in Chile (page 4, page 6, second column of page 8, page 9)\n\nB. I would develop a small taxonomy of OA policies and varieties and explain them. Not all readers are OA experts. If possible, explain relations among them, etc. Something like the description you did for Latin American OA initiatives.\nC: Revise repetitions throughout the paper. Example: In page 3:\nfirst paragraph of 1st column is repeated as fourth in second column\n\nlast paragraph first column repeated in first three in second column\n\nD. The readers of the paper are not only Chileans. Hence, do not write \"at national level\" but \"at Chilean level\", etc.\nE. Check details of form:\nbe consistent when enumerating (capital o not capital letters)\n\nthere is a [?] reference in page 9\n\nsubtitles require to be more systematic\n\nI approve this paper under the condition that the authors follow the corrections indicated.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes", "responses": [] }, { "id": "53090", "date": "28 Oct 2019", "name": "Bianca Kramer", "expertise": [ "Reviewer Expertise scholarly communication", "open science and scholarship" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article presents a view on open access and the role of repositories from a Latin American perspective, focusing on open access not only as a way to make scientific and scholarly publications more accessible, but also as part of aiming for more reproducible research.\n\nIt offers a discussion on goals of OA and reproducibility and places this in the context of institutional and government policies and practices, and the availability and usage of infrastructure in the context of Latin America, and Chili specifically.\n\nAs such, it provides an important point of view. I especially commend the authors for including Spanish-language literature in addition to English-language references.\nI agree with the first reviewer the paper could benefit from a tighter structure, removing some duplicate paragraphs, and fixing some missing and duplicate references.\nRegarding structure, I would suggest to make the aims of open access and reproducibility (in the context of Latin America) central to the paper, introducing them at the start and placing the discussion on OA definitions, policies, available infrastructure and recommendations in that context. But this is just one suggestion, the structure suggested by the first reviewer is another one.\n\nThe paper mentions the relevance of OA and reproducibility for both natural and social sciences, as well as humanities. Building on that and using the references already provided, some discussion on the differences between these main disciplines in how these topics are considered could be a useful addition to the paper.\n\nBeyond these main points, I have made a detailed annotation of the paper, including comments and suggestions (without the expectation they should all be met), as well as flagging some copy-editing issues using Hypothes.is. These are accessible here: https://hypothes.is/groups/4XjjYbQ1/f1000research-review or here.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1517
https://f1000research.com/articles/8-289/v1	14 Mar 19	{ "type": "Research Article", "title": "Boosting diabetes and pre-diabetes screening in rural Ghana", "authors": [ "Bernard Effah Nyarko", "Rosemary Serwah Amoah", "Alessandro Crimi", "Bernard Effah Nyarko", "Rosemary Serwah Amoah" ], "abstract": "Background: Diabetes is a growing worldwide disease with serious consequences to health and with a high financial burden. Ghana is one of the developing African countries where the prevalence of diabetes is increasing. Moreover, many cases remain undiagnosed, when along with pre-diabetic cases they can be easily detected. Pre-diabetes condition occurs when blood sugar levels are higher than normal but are not high enough to be classified as diabetes, and it is still reversible. The main objective of this study is to propose a novel method to increase diabetes and pre-diabetes early detection in rural areas. A secondary aim is to look for new related behavioral determinants specific to rural Ghana, by comparing subjects at risk with those already diagnosed as diabetic. Methods: The screening approach was based on tests performed pro-actively by community nurses using glucometers and mobile phone apps. As a pilot for future policies, glycemic tests were carried out on 101 subjects from rural communities in Ghana deemed at risk and unaware of their diabetic/pre-diabetic status. A comparison of dietary and lifestyle habits of the screened people was conducted in regards to a cohort of 103 diabetic patients from the same rural communities. Results: The pilot screening detected 2 diabetic subjects (2% of the cohort) showing WHO diabetic glycemic values, and 20 pre-diabetic subjects (19.8% of the cohort) which showed the effectiveness of the user-friendly approach. The need for further campaigns on alcohol consumption and physical activity has emerged, even in rural areas. Conclusions: Policies based on prevention screening as reported in the manuscript have the potential to reduce diabetes incidence, if actions are taken while patients are pre-diabetic, reduce complication related to late diagnosis and indirectly related health-care costs in the country.", "keywords": [ "diabetes", "mHealth", "community health workers", "pre-diabetes", "Ghana", "rural health" ], "content": "List of abbreviations\n\nIDF International Diabetes Federation\n\nWHO World Health Organization\n\nCHPS Community-based health planning and services\n\nBMI body mass index\n\nCDP confirmed diabetic participants\n\nUSP unknown diabetic status participant\n\n\nIntroduction\n\nDiabetes is one of the fastest growing non-communicable killer diseases in the world, claiming one life every eight seconds and a limb every 30 seconds1. Diabetes of all types can lead to complication in many parts of the body and can increase the overall risk of dying prematurely2–7. Pre-diabetes condition occurs when blood sugar levels are higher than normal, but are not high enough to be classified as diabetes; this often has no symptoms, and is reversible3.\n\nAccording to the latest 2016 data from the World Health Organization (WHO), amongst adults living with diabetes melilites, 80% live in low and middle-income countries such as those in the Asia and Eastern Pacific region. The largest number has been reported in China (90 million people6), followed by India (61.3 million people) and Bangladesh (8.4 million people)5. Complications of diabetes results in increased morbidity, disability, and mortality and have a high economic cost, especially in developing countries8. More specifically, the reported prevalence of type 2 diabetes ranges from 1% in rural Uganda to 12% in urban Kenya. While gestational diabetes has been reported in Sub-Saharan countries at varying levels (e.g. from 0% in Tanzania to 9% in Ethiopia9). Lastly, even considering those values an underestimate, it is expected that the reported cases will reach 82 million by the 203010.\n\nGhana is challenged with the increasing prevalence of diabetes, similar to other African countries, with a prevalence of 3.6% in adults and 518,000 diagnosed cases within the country11. More specifically, The prevalence of diabetes in some parts of Ghana has been found to be higher than the world average of 6.4%12,13. Moreover, the 2015 report of the IDF indicated many other cases probably remain undiagnosed, posing an increased danger of complications for people living with diabetes unaware of the consequences. Previous studies in the country showed that low level of physical activity and obesity were associated with increased risk of diabetes4. Additionally, old age and level of education were also associated with increased risk of diabetes4. It has also been observed that within Ghana sugary drinks consumption is linked to type 2 diabetes14. Community-based health planning and services (CHPS) is a national health program in Ghana adopted in 1999 to reduce geographical barriers to health care access15. According to the CHPS policy, relocating nurses directly to communities could outperform an entire sub-district health center. The cost-effectiveness of CHPS for malaria, diarrhea, and pneumonia has been recently reported16. However, specific interventions for non-communicable diseases such as diabetes have not yet been investigated within the CHPS policy.\n\nVulnerable populations such as those in low- and middle-income countries are generally more affected by diabetes related complications17. As in several fields of healthcare, mobile health (mHealth) has the potential of reducing for vulnerable populations with diabetes18. This can occur either by sending reminders or by increasing access to patient management19. Despite the plethora of studies on mHealth and diabetes management19–21, no study has been carried out in rural Africa with the aim of improving detection of diabetes and pre-diabetes by using mobile technologies and community nurses.\n\nThis study proposes a novel screening approach based on community nurses using glucometers and mobile phones, performing tests on undiagnosed and diagnosed subjects proactively within the community without waiting for participants to present at the clinic. The main objective is to develop a novel method to increase diabetes and pre-diabetes detection, and to find new behavioral determinants related to those conditions. In particular, the purpose of the mobile app is to simplify the tracking operations of the nurses and to collate the data into a centralized secure server. For this purpose, a pilot project was carried out in rural communities of the Central Region of Ghana to assess the feasibility of the approach. A secondary objective was to look for new behavioral determinants related to the rural Ghanaian populations by carrying out a comparison with a group of subjects diagnosed with diabetes from the same communities. Similar to a project carried out in the same area about improving prenatal care22, the project utilized community nurses instead of the participants to assure reliable glucose level data collection.\n\n\nMethods\n\nWe performed a community-based cross-sectional study using mixed methods of quantitative and qualitative analysis.\n\nData were collected by community nurses by using a mobile phone application and sent to a secure database. The inclusion criteria for the participants were that they were members of the study communities, the exclusion criteria included being <18 years and having a prior diagnosis of diabetes of any type. A proportional comparison group with diabetes was also recruited, with the aim of possibly finding common dietary habits with the screened subjects found to be diabetic/pre-diabetics in both groups. The inclusion criteria for the comparison group were having a prior diagnosis of type 2 diabetes, being part of the rural communities sampled, and being older than 18 years. To avoid unnecessary overtime for the nurses, we excluded children and young adults under the age of 18 years; we acknowledge this as limitation for our study.\n\nSubjects were assessed as “non-diabetic” according to their diabetes status awareness, obtained by the following question: \"Has a doctor or another health professional ever diagnosed you have diabetes?\" Further variables analyzed were: family history (family member diagnosed with diabetes); pregnancy; history of hypertension; screened glucose level; lifestyle characteristics (going to sleep within 1 hour after dinner and level of physical activities); body mass index (BMI), and diet (consumption per typical week of dishes based on staple or maize/corn, root and tuber-potatoes/cassava, and alcohol). Those variables are further described in the following sections as whether they were assessed by the nurses or self-reported.\n\nThe research was conducted in the central region of Ghana, specifically within the Biriwa and Anomabo communities. These two communities are in the Mfantsiman Municipal District and based on the 2010 census, the two towns have a population of about 7,500 and 14,389 respectively. Those communities have been chosen due to them being relatively close to a main road connecting urban centers. The assumption is that the members of those communities are more prone to adopt unhealthy habits (such as smoking and drinking) which are more common in urban centers.\n\nCommunity nurses from rural clinics were instructed to visit rural communities performing glucose screening, when fasted if possible or alternatively at random. Those tested were subjects known to have diabetes, subjects deemed at risk or those willing to be tested. A total of 204 people were tested in a window period of 6 months (from June to December 2017). This sample size was reached following the minimum sample size for the study given by two populations of n=86 subjects. This minimum sample size was computed by using the GPower software version 3.1 for an a priori two-tail t-test with alpha = 0.05, effect size 0.5 and power 0.90. The quantified sample size also matches constraints according to logistics related to the community nurses. Indeed, the aim was for the proposed screening to be performed by the nurses in addition to their normal activities without resulting in the need for overtime or compromising the other activities were carrying out. As one community nurse per community was used, in this pilot two nurses were employed.\n\nThe nurses were equipped with glucometers and low-cost Android smart phones. They received a short (less than one day) training on the app and were supported on its use during the first week of the project. Data were stored through the mobile phone app and sent to a server to improve management and facilitate eventual longitudinal screening. The developed app was based on the CommcareHQ framework, and comprised a series of guided questions that the nurses completed in addition to the glucose test (questions used are available as Extended data23). Some screen-shots of those questions are depicted in Figure 1. If network was not available at the point of data collection, information could be sent later when the network was available. CommcareHQ is a popular mobile data collection, and it has been used in several projects. For a review on projects using the CommcareHQ platform, the reader is addressed to 24.\n\nThe questions related to diet were based on general consumption for a typical week and not the week at point of sampling.\n\nFigure 2 shows the typical two steps of the screening, first a nurse is performing a glucose test (on the left), and then the data are recorded through the app (on the right). Collected data from respondents were from the rural communities, both male and female, including pregnant women with the intent of capturing eventual gestational diabetes case25. The sampling of candidates at risk was based on physical factors and at the nurse’s discretion. During regular visits to the rural communities the nurses assessed whether the person to be tested was a potential diabetic candidate and therefore deemed to be tested (e.g. appearing overweight), paying particular attention to pregnant women. By doing this we have introduced a piloted bias (as overweight people, pregnant women and other cases were deemed at risk by the nurses). Therefore, with those biases, the resulting statistics is not representative of a random screening of the national population.\n\nNevertheless, our focus is not to estimate total incidence of undetected diabetes or pre-diabetes, but to propose an approach that can detect as many as possible cases which otherwise would go unnoticed.\n\nHowever, to avoid a strong bias individual deemed healthy that were willing to be tested were included. Any detected diabetic or pre-diabetic cases were immediately informed, and lifestyle changes or pharmaceutical therapies discussed. Data were collected on known diabetic patients as well, to evaluate the differences or similarities in lifestyle among the various groups.\n\nWith the mobile phone app, the community nurses could also keep track of longitudinal changes or whether a subject has been already tested, in a similar manner of a project conducted in the same area about boosting prenatal care22. Subjects who came into contact with community nurses were asked the last time they had a meal and based on the response blood glucose measurements were classed as either fasting blood sugar levels or random blood sugar levels. Furthermore, the following information were recorded from the subject a) Anthropometric indices (weight, height, BMI) measured by the nurses b) Demographic Information (sex, age) self-reported c) Blood glucose measurement (fasting or random sugar test) measured by the nurses d) Information on risk factors (pregnancy, family history of diabetes, history of hypertension, kidney disorder, alcoholism, low levels of exercise and unhealthy eating habits) self-reported. OneTouch Ultra (Milpitas, CA, USA) and Accu-Check (Hoffmann–La Roche SA, Basel, Switzerland) strips and glucometers were used to measure the blood glucose level. According to the WHO classification26, a fasting blood glucose level greater 110 mg/dL but less than 126 mg/dL was considered pre-diabetic. A fasting blood sugar level 126 mg/dL or above was considered diabetic. A random blood sugar level greater 140 mg/dL but less 200 mg/dL was considered pre-diabetic. Above 200mg/dL was considered diabetic. Family history of diabetes is referred to occurrence of diabetes in close relatives defined as either father, mother, siblings or offspring. Most of the data were assessed by the nurse though some were self-reported by the subjects and this might represent a limitation. The data entered through the developed Android app were stored in a secure server provided by Dimagi. Data were entered in the forms of the mobile app and transmitted, encrypted, to the cloud-based server, where they were accessed and downloaded via a password-protected web interface.\n\nThe main aim of the study was to determine if it is possible to easily detect diabetic and pre-diabetic subjects through community nurses already involved and active within the CHPS policy. Additionally, dietary habits and demographic information were collected by the community nurses along with glucose level to explore novel determinants related to the disease, which have not been reported in literature so far.\n\nThe data entered through the mobile phone app were downloaded and analyzed by R computing software version 3.5.2. Quality check on the data was performed and forms deemed clearly erroneous were removed. Statistically significant relationships among the collected information were sought comparing the population at risk and the population with assessed diabetes by using a two-tail t-test, and the related p-value less or equal to 0.05 was considered to indicate a significant value.\n\nThe study also took advantage of the screening process to collect further insights through qualitative methods. Hence, the quantitative information was complemented with qualitative data obtained through semi-structured interviews performed by the authors with the community nurses to identify further relevant elements at the end of the pilot project. After revision of notes, the transcripts were typed and coded by using NVivo 10 (QSR International, Melbourne, Australia). The interviews were thus analyzed by using qualitative conventional content analysis. The starting open questions were “what are your general comments about the projects?”, “Which shortcomings did you notice?”, “What are your suggestions?”.\n\n\nResults\n\nThe diabetes screening saw over 204 inhabitants in Anomabo and Biriwa over a period of 6 months. Of those, 103 were previously confirmed diabetic participant (CDP) with an average age of 62.9 ±11.2 years, and 101 people, with an average age of 30 ±9.7 years, with unknown diabetic status participant (USP). Originally 211 forms were completed, however, 7 of them were deemed erroneous during the quality check and therefore were removed before the analysis. For each person data were collected only once. The CDP cohort comprised 66 female and 37 male subjects, while the USP cohort comprised 95 female and 6 male subjects (see Underlying data23). Details of the main demographic characteristics revealed for the two groups are reported in Table 1 while the breakdown of the study variables is distributed in the other tables . Community nurses see on average 20 patients per day 5 times a week. Their duties mainly encompass malaria, diarrhea, and pneumonia treatments which are generally perceived as more urgent. During the pilot, the total number of people approached for the diabetes screening was 240 with a participation rate of 88%. The 12% of who refused the screening reported as the main motivation the unwillingness to sign the written consent for the study.\n\nTwo-sample t-test performed comparing the BMIs was statistically significant (P value <0.05) although they were both normoweight. Some subjects of the CDP cohort also presented with co-occurrence of ulcers (n=4), asthma (n=1), arthritis or rheumatism (n=3), and kidney disease (n=1). At time of testing participants in the USP cohort presented with co-occurrence of asthma (n=4), arthritis and rheumatism (n=1), and typhoid fever (n=1). The subject presenting typhoid fever showed a random glucose blood level of 132 mg/dL which could not be considered neither diabetic nor pre-diabetic. Therefore, typhoid fever was not considered a confounding factor as the tested participant did not show a high value due to this. These results are summarized in Table 2.\n\ndemographic characteristics for the cohort. BMI is reported as means and standard deviations, (co-) occurrence of hypertension or close relative with diabetes, while diseases are reported in absolute values.\n\nDuring the proactive screening performed by the community nurses, two subjects (1 female, not pregnant, with hypertension, 35.4 BMI; 1 male, 25.7 BMI) were found to be hyperglycemic at fasting which would diagnose them as diabetic according to the current WHO threshold26. These subjects were not aware of their condition despite close relatives with diagnosed diabetes (son in one case and siblings in the other). They did not present any further symptoms, and they were not habitual consumers of alcohol or red meat. However, their diet was based on dishes with large amounts of maize corn, cassava and rice.\n\nIn total, 20 pre-diabetic cases were identified according to the WHO threshold, four tested at fasting and 16 at random (19 female; 1 male). No hypertension or other symptoms were identified, and 5 of had relatives with diabetes (mother or father). Half of the detected subjects reported consuming red meat almost daily, and all claimed to avoid alcohol consumption. They also reported frequently consuming dishes comprising of maize corn, cassava and rice, averaging respectively 8.5, 8.5 and 3.5 times per week.\n\nAll subjects of both groups claimed they were used to performing physical activities due to their daily job. No statistical difference (p-value >0.05) across the two cohorts was detected regarding weekly consumption of red meat, maize corn, cassava or rice. The mean consumption of alcohol across the two populations was also not significantly different (p-value >0.05), however, in the CDP cohort 14 subjects declared to consume typically at least 1 alcoholic beverage per week (plus 10 claimed to be former alcohol drinkers before their diagnosis and then changed this behavior) while in the USP cohort only 6 participants indicated they consumed at least 1 alcoholic beverage per week. Furthermore, in the CDP cohort 3 subjects claimed to have reduced the consumption of cassava based products, and another 2 for red meat. 88 subjects reported to have drastically reduced the consumption of sugar, salt or both but not to have altered their diet. No case of gestational diabetes was detected. Table 3 reports the mean and standard deviation of the blood glucose level for both cohorts, distinguishing whether tests were fasting or random. Summarizing qualitatively, no specific novel determinant was found.\n\nThe two nurses who performed the screening in the rural areas were asked to give their qualitative opinion at the end of the 6 months pilot. The following are the main extracts of those interviews.\n\n- Enrolled community nurse, Anomabo Health Centre, Anomabo:\n\n“There should be continuous education of the masses on diabetes to create awareness. There should also be financial support from governments, and non-governmental organizations to aid routine check up of known diabetes patients, this will encourage them to always take their medication as regular checking of blood glucose level for free help them to know the progress with their condition.”\n\n“Medication as insulin is covered by national insurance but not the needles for the screening, and this can make people refrain from screening.”\n\n- Enrolled community nurse, Boabab Health Clinic, Biriwa:\n\n“Most patients were willing to be tested and ready to give any information on their personal lifestyle, diet and medication. There is a general trust in tests performed by clinical personnel regardless on cultural beliefs and on the fact that we are a small clinic.”\n\n“Particularly, female clients above the age of forty were pleased to participate in the screening exercise.”\n\n“Furthermore, some clients were not able to provide precise information about their diet and family history.”\n\n“Screened people reported performing some kind of sport activity related to their job, believing it was sufficient to keep them healthy.”\n\n\nDiscussions\n\nMobile phone apps and pro-active screening can help community nurses to spot new cases of diabetes and pre-diabetes. In particular, the proposed approach was based on pro-actively performing blood screening during rural visits of the community nurses, who were collecting information via the mobile phone app. This can also help tracking and monitoring, as the nurses can follow up the status of the participants during future visits.\n\nOur findings indicate that some individuals in vulnerable populations, such as those in rural Ghana, are not aware of becoming diabetic or being in diabetic condition. We report two cases of diabetic participants in the USP cohort (2%) and 20 as pre-diabetic (19.8%), which we consider to be high when compared to previously reported statistics27. However, it must be taken into account that the cohort selection was not purely random, and piloted bias was introduced as the subjects to be tested were chosen according to the nurse’s discretion. Therefore, these percentages should not be taken as a representative sample of the national population. The proposed approach aims instead at detecting as many cases as possible of diabetics and pre-diabetics which otherwise would go unnoticed. For this purpose it proved to be successful, inexpensive and easily integrated into the standard duties of community nurses.\n\nInitially the nurses were equipped with ihealth glucometer dongles for the smartphone (ihealth, Mountain View, CA, USA). The anecdotal comments from nurses were that. despite the initial interest, they were not practical to collect data, though they might be suitable for a single individual. The reason can be related to the familiarity of the nurses to known tools such as Accu-Check and OneTouch, or the cumbersome use of switching continuously between the mobile app of the glucometer dongle and the app to record the data. The use of a mobile app can allow for large institutions to easily monitor diabetes in rural areas through the collected data held in a secure centralized server. Moreover, mobile apps appear more user friendly for the nurse in comparison to cumbersome paperwork.\n\nDuring the qualitative interviews, one nurse pointed out that despite the marginal costs of glucose tests both patients and government are not promoting them, while it could be cost-effective for Ghanaian institutions to detect pre-diabetic cases instead of dealing with a growing diabetic population, as has been shown for similar vulnerable populations28. Despite the progresses made in Ghana to achieve universal health access and coverage, financial barriers to diabetes service utilization still exist. Subjects usually do not undergo glucose tests due to financial constraints, as the test is not covered under most standard medical insurance policies, as is the case in Nigeria and Tanzania29. One nurse mentioned that subjects were often not able to report their dietary habits and were unaware of the effects of their diet on their health. Conversely to prenatal care30, it appears that generally the members of the communities trust the personnel of the clinics for this type of tests.\n\nAt population level the CDP and USP cohort were normoweight (having a BMI between 18.5 and 24,99), though there were subjects which were obese and with hypertension in both groups. It is worthwhile to mention that the CDP subjects might have some dietary changes already after being informed of being diabetic, which could have affected their weight and other measurements. However, from the interviews it seems that the major changes were the reduction of consumption of salt and sugar, some increased their consumption of vegetables and fruit, and some reducing the consumption of alcohol beverages. The study did not track behavioral changes, but we can presume that these may have occurred and can represent confounding factor. No statistically significant difference in alcohol consumption between the two groups was detected. However, this could be due to the fact that some individuals in the CDP cohort changed their dietary habits (10 people reported to have cut out alcohol consumption after their diagnosis). Moreover, the nurses were instructed to focus for both groups on women which might consume fewer alcoholic beverages than men. Nevertheless, the general impression of the nurses was that the alcohol consumption had an impact. With the increase in quality of life in the country, western habits such as alcohol consumption might be also increasing, and therefore augmenting the risk of diabetes. It has been noticed that people informed of their condition tend to change their dietary habits. Nevertheless, given the country-wide growing trend in alcohol consumption, social marketing campaign related to this 31 should be considered.\n\nAlmost all subjects of both groups reported to perform regular physical activities related to their job. However, this information seems vague and it is not clear whether this physical activity is aerobic or resistance or whether it is sufficient to keep a healthy glucose level. Most likely further activities should be proposed. Jogging and other sport activities are inexpensive and easy to promote. Therefore, promoting this type of sport activities can address this issue. Moreover, it appears necessary in the future to use more detailed investigations about sport activities, such as using the WHO global physical activity questionnaire32.\n\nGhana has experienced an exponential increase of the mobile network, social media and smartphones in the recent years33. Beyond the screening of the population carried out by nurses, smartphones can have an impact on glycemic control, as smartphone dongles can be inexpensive and attractive to young users. Strategies such as gamification, and social media should be explored to increase awareness on glycemic levels as shown in other contexts34.\n\n\nConclusions\n\nProactive glycemic screening on vulnerable population – such as those living in rural areas – can be effective in detecting new cases of diabetes and pre-diabetes. Our approach using community nurses screening subjects deemed at risk and collecting data on mobile phone was found to be effective, and suitable for longitudinal studies. Campaigns increasing awareness of alcohol consumption, physical activity, nutrition and healthy habits should be emphasized in any prevention strategy as the population seems to still be unaware of the consequences.\n\nDespite this, studies with larger population are required to confirm the results. The diabetes and pre-diabetes screening described in this manuscript can be easily included into the national CHPS policy with several potential benefits. Those benefits include reducing incidence by detecting cases of pre-diabetes which hopefully will not convert into type-2 diabetes and enabling timely treatment of diabetes patients avoiding complications related to delays in treatment.\n\n\nEthics and consent\n\nAll procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Written consent for the reported data was collected. The Noguchi Memorial Institute for Medical Research of the University of Ghana recorded the study with the identifier 076/13-14.\n\n\nData availability\n\nZenodo: DiabetesUP: Initial repository of data related to Nyarko et al. http://doi.org/10.5281/zenodo.258711723\n\nThis project contains the following underlying data:\n\nDiabetes_cohort.xls (Collected data for diabetes cohort)\n\nUnaware_cohort.xls (Collect data for participants with no prior diagnosis of diabetes)\n\nZenodo: DiabetesUP: Initial repository of data related to Nyarko et al. http://doi.org/10.5281/zenodo.258711723\n\nThis project contains the following extended data:\n\nQuestionnaires.odt (Questions asked by the app)\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).\n\n\nSoftware availability\n\nThe CommCare software is required to use the source code, questions used in the app are available as Extended data\n\nSoftware available from: https://play.google.com/store/apps/details?id=org.commcare.dalvik&hl=en\n\nSource code available from: https://github.com/alecrimi/diabetesUP/tree/v1.0\n\nArchived source code at time of publication: http://doi.org/10.5281/zenodo.258711723\n\nLicense: Creative Commons Zero \"No rights reserved\" data waiver", "appendix": "Grant information\n\nThis study was partially funded by the Regional Registry for Internet Number Resources serving the African Internet Community (Afrinic).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThis research was conducted with the support of Baobab medical center in Biriwa (Ghana) and Anomabo Health Center (Ghana) and the related communities. We are thankful to Emma K. Capodaglio for kindly copyedit the paper.\n\n\nReferences\n\nZhang Y, Hu G, Zhang L, et al.: A novel testing model for opportunistic screening of pre-diabetes and diabetes among U.S. adults. PLoS One. 2015; 10(3): e0120382. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDeshpande AD, Harris-Hayes M, Schootman M: Epidemiology of diabetes and diabetes-related complications. Phys Ther. 2008; 88(11): 1254–1264. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTabák AG, Herder C, Rathmann W, et al.: Prediabetes: a high-risk state for diabetes development. Lancet. 2012; 379(9833): 2279–2290. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGatimu SM, Milimo BW, Sebastian MS: Prevalence and determinants of diabetes among older adults in Ghana. BMC Public Health. 2016; 16(1): 1174. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAkter S, Rahman MM, Abe SK, et al.: Prevalence of diabetes and prediabetes and their risk factors among Bangladeshi adults: a nationwide survey. Bull World Health Organ. 2014; 92(3): 204–13, 213A. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhang P, Gregg E: Global economic burden of diabetes and its implications. Lancet Diabetes Endocrinol. 2017; 5(6): 404–405. PubMed Abstract \| Publisher Full Text\n\nHippisley-Cox J, Coupland C: Diabetes treatments and risk of amputation, blindness, severe kidney failure, hyperglycaemia, and hypoglycaemia: open cohort study in primary care. BMJ. 2016; 352: i1450. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPapatheodorou K, Papanas N, Banach M, et al.: Complications of Diabetes 2016. J Diabetes Res. 2016; 2016: 6989453. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHall V, Thomsen RW, Henriksen O, et al.: Diabetes in Sub Saharan Africa 1999-2011: epidemiology and public health implications. A systematic review. BMC Public Health. 2011; 11(1): 564. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWild SH, Roglic G, Green A, et al.: Global prevalence of diabetes: estimates for the year 2000 and projections for 2030: response to Rathman and Giani. Diabetes care. 2004; 27(10): 2569. Publisher Full Text\n\nInternational Diabetes Federation, Diabetes Atlas Eighth edition. 2017. Reference Source\n\nde Graft Aikins A, Owusu-Dabo E, Agyemang C: Diabetes in Ghana: a review of research on prevalence, experiences and healthcare. Chronic Noncommunicable Diseases in Ghana: Multidisciplinary Perspectives. 2014; 1: 41. Reference Source\n\nAmoah AG, Owusu SK, Adjei S: Diabetes in Ghana: a community based prevalence study in Greater Accra. Diabetes Res Clin Pract. 2002; 56(3): 197–205. PubMed Abstract \| Publisher Full Text\n\nBuxton C, Hagan JE: A survey of energy drinks consumption practices among student -athletes in Ghana: lessons for developing health education intervention programmes. J Int Soc Sports Nutr. 2012; 9(1): 9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNyonator FK, Awoonor-Williams JK, Phillips JF, et al.: The Ghana community-based health planning and services initiative for scaling up service delivery innovation. Health Policy Plan. 2005; 20(1): 25–34. PubMed Abstract \| Publisher Full Text\n\nEscribano Ferrer B, Hansen KS, Gyapong M, et al.: Cost-effectiveness analysis of the national implementation of integrated community case management and community-based health planning and services in Ghana for the treatment of malaria, diarrhoea and pneumonia. Malar J. 2017; 16(1): 277. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPeek ME: Can mHealth Interventions Reduce Health Disparities among Vulnerable Populations? Divers Equal Health Care. 2017; 14(2): 44–45. Publisher Full Text\n\nOsborn CY, Mayberry LS, Peek ME, et al.: Recent mHealth and Internet interventions for disadvantaged and vulnerable people with type 2 diabetes: Considerations and recommendations for future research. Diversity in Health and care. (2017), 2014; 14(2): 44–45.\n\nKitsiou S, Paré G, Jaana M, et al.: Effectiveness of mHealth interventions for patients with diabetes: An overview of systematic reviews. PLoS One. 2017; 12(3): e0173160. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHall AK, Cole-Lewis H, Bernhardt JM: Mobile text messaging for health: a systematic review of reviews. Annu Rev Public Health. 2015; 36: 393–415. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBaron J, McBain H, Newman S: The impact of mobile monitoring technologies on glycosylated hemoglobin in diabetes: a systematic review. J Diabetes Sci Technol. 2012; 6(5): 1185–96. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAmoah B, Anto EA, Osei PK, et al.: Boosting antenatal care attendance and number of hospital deliveries among pregnant women in rural communities: a community initiative in Ghana based on mobile phones applications and portable ultrasound scans. BMC Pregnancy Childbirth. 2016; 16(1): 141. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCrimi A: DiabetesUP: Initial repository of data related to Nyarko et al. (Version v1.0). Zenodo. 2019. http://www.doi.org/10.5281/zenodo.2587117\n\nMhila G, DeRenzi B, Mushi C, et al.: Using mobile applications for community-based social support for chronic patients. Health Informatics in Africa. 2009. Reference Source\n\nCrowther CA, Hiller JE, Moss JR, et al.: Effect of treatment of gestational diabetes mellitus on pregnancy outcomes. N Engl J Med. 2005; 352(24): 2477–2486. PubMed Abstract \| Publisher Full Text\n\nWorld Health Organization: Definition and diagnosis of diabetes mellitus and intermediate hyperglycemia: report of a WHO/IDF consultation. World Hearth Org. 2006. Reference Source\n\nCook-Huynh M, Ansong D, Steckelberg RC, et al.: Prevalence of hypertension and diabetes mellitus in adults from a rural community in Ghana. Ethn Dis. 2012; 22(3): 347–52. PubMed Abstract\n\nMash R, Kroukamp R, Gaziano T, et al.: Cost-effectiveness of a diabetes group education program delivered by health promoters with a guiding style in underserved communities in Cape Town, South Africa. Patient Educ Couns. 2015; 98(5): 622–626. PubMed Abstract \| Publisher Full Text\n\nMwangome M, Geubbels E, Klatser P, et al.: Perceptions on diabetes care provision among health providers in rural Tanzania: a qualitative study. Health Policy Plan. 2017; 32(3): 418–429. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGanle JK: Ethnic disparities in utilisation of maternal health care services in Ghana: evidence from the 2007 Ghana Maternal Health Survey. Ethn Health. 2016; 21(1): 85–101. PubMed Abstract \| Publisher Full Text\n\nKotler P, Zaltman G: Social marketing: an approach to planned social change. J Mark. 1971; 35(3): 3–12. PubMed Abstract \| Publisher Full Text\n\nGlobal physical activity questionnaire: Department of chronic diseases and Health promotion surveillance and population based prevention. World Health Organization, Geneva, Switzerland. Reference Source\n\nAsongu SA, Le Roux S: Enhancing ICT for inclusive human development in Sub-Saharan Africa. Technol Forecast Soc Change. 2017; 118: 44–54. Publisher Full Text\n\nCafazzo JA, Casselman M, Hamming N, et al.: Design of an mHealth app for the self-management of adolescent type 1 diabetes: a pilot study. J Med Internet Res. 2012; 14(3): e70. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "49360", "date": "25 Jun 2019", "name": "Juan Salazar", "expertise": [ "Reviewer Expertise Internal medicine", "Diabetes. Epidemiology", "Chronic disease" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a report by Nyarko et al., who propose a novel screening strategy for diabetes mellitus in a rural population of Ghana, Africa. An aspect of regional epidemiological relevance that could be reproduced in other geographic locations with similar socioeconomic characteristics.\n\nAbstract:\nIntroduction: it is not necessary to define prediabetes in this line.\n\nIn Methods, it should describe the type of study and sampling. How were those 101 subjects chosen?\n\nIn Results, the sentence \"The need for further campaigns on alcohol and physical activity has emerged, even in rural areas\", must go in Conclusions not in Results.\n\nIntroduction:\n\nIt is somewhat extensive, paragraphs 2 and 3 could be more concise, emphasizing the justification for the design of this study.\n\nMaterials and methods:\n\nThis section is somewhat disorganized, especially at the sample selection section. For this, it would be convenient to divide by subtitles according to the stages of the project, e.g.:\n\n- Study design and sample selection\n\n- Screening phase\n\n- Diabetes subjects selection phase\n\n- Clinical and capilary sugar evaluation\n\n- Definitions\n\n- Data analysis\n\n- Quantitative approach\n\n- Qualitative approach\nYou must specify what type of sampling was used for both groups analyzed.\n\nWere the questions used by the mobile app validated or standardized, and by what method? There are some standardized questionnaires to assess the risk of diabetes in \"healthy\" subjects, such as the Diabetes Risk Test (ADA) and FINDRISK test. Why were some of these not used?\n\nResults:\nThere are some characteristics that must be specified in a table or figure:\n\nIn the USP group, how many subjects were prediabetic and diabetic according to the method used?\n\nWhat were the determinants or risk factors most often evidenced in this group?\n\nIn the CDP group, how many subjects received pharmacological treatment? What were the most commonly used drugs? What were the factors or determinants of risk in control disease?\n\nAlthough some comparisons are described in the text, as a screening study, some characteristics must also be specified in a table or figure to evaluate the proportions comparatively. In table 2 it is important to include the variables related to psychobiological habits, family and personal history of cardiovascular diseases and their risk factors.\n\nDiscussion:\nIt would also be advisable to organize this section (without the need to place subtitles) to make it easier to read.\n\nWhat is known about this topic? Especially in your region. Main findings of the study.\n\nWhat does this study add?\n\nLimitations: Mention in detail in this section.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4847", "date": "27 Aug 2019", "name": "Alessandro Crimi", "role": "Author Response", "response": "We have listed here our answers starting with an «(A)».Abstract: Introduction: it is not necessary to define prediabetes in this line. (A) We have removed the sentence. In Methods, it should describe the type of study and sampling. How were those 101 subjects chosen? (A) As we explained later in the paper, subjects were chosen through snowball sampling. In Results, the sentence \"The need for further campaigns on alcohol and physical activity has emerged, even in rural areas\", must go in Conclusions not in Results. (A) We agree with the reviewer. Introduction: It is somewhat extensive, paragraphs 2 and 3 could be more concise, emphasizing the justification for the design of this study. (A) We relatively disagree with the reviewer. Paragraph 2 is about Diabetes in low and middle income countries, while Paragraph 3 is about Ghana. We do believe all information reported is currently relevant. Otherwise we ask the reviewer to point out which sentences are not relevant.Materials and methods: This section is somewhat disorganized, especially at the sample selection section. For this, it would be convenient to divide by subtitles according to the stages of the project, e.g.: - Study design and sample selection - Screening phase - Diabetes subjects selection phase - Clinical and capilary sugar evaluation - Definitions - Data analysis - Quantitative approach - Qualitative approach(A) We have inserted further subtitles suggested by the reviewer and shuffled some sentences to fit the structure. You must specify what type of sampling was used for both groups analyzed. (A) Snow-balling sampling entrusted to the community nurses was used. Were the questions used by the mobile app validated or standardized, and by what method? There are some standardized questionnaires to assess the risk of diabetes in \"healthy\" subjects, such as the Diabetes Risk Test (ADA) and FINDRISK test. Why were some of these not used? (A) The authors are aware of existing standardized questionnaires. However, they are relatively simplistic and comprising basic information not specific for the case in exam. Indeed most of the questions of the standardized questionnaires are included in our ad-hoc questionnaire. We envision our questionnaire as an expansion of the standardized questionnaire. Nevertheless, we are including this clarification. Results: There are some characteristics that must be specified in a table or figure: In the USP group, how many subjects were prediabetic and diabetic according to the method used? (A) This is information is already included in the paper. This is even mentioned in the abstract. What were the determinants or risk factors most often evidenced in this group? (A) Unfortunately we have not found novel strong dietary determinants. The existence of relatives with diabetes for the pre-diabetic group is not a novel discovery. Our contribution is that it is possible to perform early detection of diabetes and pre-diabetes and that there is still a general misconception of what «physical activity» is. In the CDP group, how many subjects received pharmacological treatment? What were the most commonly used drugs? What were the factors or determinants of risk in control disease? (A) All subjects in the CDP group were following pharmacological treatment, but we do not know the details of this. Although some comparisons are described in the text, as a screening study, some characteristics must also be specified in a table or figure to evaluate the proportions comparatively. In table 2 it is important to include the variables related to psychobiological habits, family and personal history of cardiovascular diseases and their risk factors. (A) We are not in possession of detailed information about cardiovascular diseases apart from hypertension. However, if not reported most likely there is no past medical history of cardiovascular diseases for the subjects in exam. The study focused on the early detection of diabetes and pre-diabetes, we will acknowledge in the discussion section that deeper investigation on cardiovascular disease is missing and represents a limitation. Discussion: It would also be advisable to organize this section (without the need to place subtitles) to make it easier to read. What is known about this topic? Especially in your region. Main findings of the study. What does this study add? Limitations: Mention in detail in this section. (A) We have restructured the discussion section, particularly adding a paragraph on «what is known» and a paragraph about «limitations»." } ] }, { "id": "50381", "date": "01 Jul 2019", "name": "Frank Peter Schelp", "expertise": [ "Reviewer Expertise Public Health", "epidemiology", "nutrition", "NCDs", "Infectious diseases", "health care systems in low- and middle income countries" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nFrom a historical point it is of interest, that this publication is submitted from authors from Ghana. That is remarkable, as the topic deals with diabetes mellitus, a disease, mainly due to overnutrition and obesity, from a country, where Cecily Delphine Williams once worked and published in 1935 the key paper about ‘Kwashiorkor’, one of the most severe and deadly conditions due to undernutrition. This reflects the drastic change of public health problems over time, due to the ‘epidemiological transition’.\nThe manuscript is also remarkable, in that it shows how possibly a community approach, might help to control diabetes mellitus. This is within the spirit of ‘primary health care’, applied to a non-communicable disease.\nThe authors report how community health nurses, equipped with inexpensive, appropriate technical innovations, measure capillary blood glucose, assess the BMI, blood pressure and collect information about additional diseases. For the study qualitative and quantitative research methods were applied. The authors repeatedly point out, that one of the major intentions of the project was to study the possibility and feasibility of using community health nurses for working against diabetes mellitus. The paper, with major improvements, will be acceptable as far as solely the qualitative approach is followed in trying to meet the objective. From the point of view of epidemiology however, the quantitative approach is inadequate. To use the term ‘screening’ is seriously misleading (please see page 257 and 259 of Porta, 20141). To ‘introduce a piloted bias’ is an innovative euphemism by the authors and it seems, that they have their own doubts about a meaningful quantitative assessment based on groups of participants being too small and too heterogeneous.\n\nThe authors might consider to change the now misleading title, to one which hints towards the objective of the publication more clearly.\n\nThe last paragraph of the introduction should be revised and should mention the aims of conducting some sort of quality study.\n\nThe methods and results sections might be combined. More inside information about the communities would be of benefit. Briefly mentioning the assumed knowledge of the participants about diabetes and the attitude towards the disease could help to estimate the usefulness of intervention. A short explanation how the community health worker fit into the overall health care for communities at the periphery of the overall health care system would be worthwhile as well. The section could be structured along the methods applied and for each method the usefulness, the acceptance of the participants within the communities, and the experience of the field worker could be described.\n\nIn the discussion the overall experience gained of the project should be summarized in view of how an attempt to control diabetes mellitus in the communities in screening for new cases and caring for diseased patients would be helpful for further investigations and will be informative for local as well as national health workers and administrators. Additional points of interest would be to briefly mention the general dietary intake of the population and how this relates to overweight and diabetes. Last but not least, the question might be addressed on whether cut-off points for ‘normal’ and ‘not-normal’ glucose levels, as suggested internationally, are meaningful for Ghana, and what is the reaction of the authors to the recent suggestion to disregard ‘prediabetes’ as a diagnosis (see Piller, 20192).\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4848", "date": "27 Aug 2019", "name": "Alessandro Crimi", "role": "Author Response", "response": "We have listed here our answers starting with an «(A)». The paper, with major improvements, will be acceptable as far as solely the qualitative approach is followed in trying to meet the objective. From the point of view of epidemiology however, the quantitative approach is inadequate. To use the term ‘screening’ is seriously misleading (please see page 257 and 259 of Porta, 20141). (A) We agree with the reviewer that given our context and sample size the term «detection» could be more suitable. To ‘introduce a piloted bias’ is an innovative euphemism by the authors and it seems, that they have their own doubts about a meaningful quantitative assessment based on groups of participants being too small and too heterogeneous. (A) The authors are sure about the approach considering the goal of increasing early detection and we trust that «our introduced piloted bias» for the purpose is sensible. However, we are perfectly aware that due to funding we are using a small sample size; this is reported more strongly in the limitation of the study. The last paragraph of the introduction should be revised and should mention the aims of conducting some sort of quality study. (A) We agree with the reviewer, and we have softened the part regarding «novel determinants» which is not the main aim of the study. The methods and results sections might be combined. More inside information about the communities would be of benefit. Briefly mentioning the assumed knowledge of the participants about diabetes and the attitude towards the disease could help to estimate the usefulness of intervention. A short explanation how the community health worker fit into the overall health care for communities at the periphery of the overall health care system would be worthwhile as well. The section could be structured along the methods applied and for each method the usefulness, the acceptance of the participants within the communities, and the experience of the field worker could be described. (A) The idea of merging methods and results could be unproductive, and actually against the comments of the other reviewer which is asking even more structure. The additional information might be useful and we acknowledge that maybe we considered it as obvious. More specifically, the communities in exam are used to interact with community nurses and there is a high level of trust. The population is also generally aware of the importance of early diagnosis, though very often they restrain from it due to the related costs not covered by the normal insurance. In the discussion the overall experience gained of the project should be summarized in view of how an attempt to control diabetes mellitus in the communities in screening for new cases and caring for diseased patients would be helpful for further investigations and will be informative for local as well as national health workers and administrators. Additional points of interest would be to briefly mention the general dietary intake of the population and how this relates to overweight and diabetes. Last but not least, the question might be addressed on whether cut-off points for ‘normal’ and ‘not-normal’ glucose levels, as suggested internationally, are meaningful for Ghana, and what is the reaction of the authors to the recent suggestion to disregard ‘prediabetes’ as a diagnosis (see Piller, 20192). 1. Porta M: A Dictionary of Epidemiology, Sixth Edition. Oxford University Press. 2014.2. Piller C: Dubious diagnosis.Science. 2019; 363 (6431): 1026-1031(A) The paper was submitted in early 2018, when the «opinion» of C. Piller was published at the beginning of 2019. We share the concerns in Europe and US about the abuse of medication for pre-diabetes. This is however, out of place especially considering the context of rural Ghana. Nevertheless, we thank the reviewer for pointing this out. We agree that we should avoid medications for pre-diabetes, while we should promote more awareness on inexpensive physical activities as jogging. Most of the points raised by the reviewers can be addressed by further investigation on physical activities and social marketing campaigns on physical activities." } ] } ]	1	https://f1000research.com/articles/8-289
https://f1000research.com/articles/8-465/v1	14 Apr 19	{ "type": "Research Article", "title": "Analysis of correlation-based biomolecular networks from different omics data by fitting stochastic block models", "authors": [ "Katharina Baum", "Jagath C. Rajapakse", "Francisco Azuaje", "Katharina Baum" ], "abstract": "Background: Biological entities such as genes, promoters, mRNA, metabolites or proteins do not act alone, but in concert in their network context. Modules, i.e., groups of nodes with similar topological properties in these networks characterize important biological functions of the underlying biomolecular system. Edges in such molecular networks represent regulatory and physical interactions, and comparing them between conditions provides valuable information on differential molecular mechanisms. However, biological data is inherently noisy and network reduction techniques can propagate errors particularly to the level of edges. We aim to improve the analysis of networks of biological molecules by deriving modules together with edge relevance estimations that are based on global network characteristics.\n\nMethods: We propose to fit the networks to stochastic block models (SBM), a method that has not yet been investigated for the analysis of biomolecular networks. This procedure both delivers modules of the networks and enables the derivation of edge confidence scores. We apply it to correlation-based networks of breast cancer data originating from high-throughput measurements of diverse molecular layers such as transcriptomics, proteomics, and metabolomics. The networks were reduced by thresholding for correlation significance or by requirements on scale-freeness.\n\nResults and discussion: We find that the networks are best represented by the hierarchical version of the SBM, and many of the predicted blocks have a biological meaning according to functional annotation. The edge confidence scores are overall in concordance with the biological evidence given by the measurements. As they are based on global network connectivity characteristics and potential hierarchies within the biomolecular networks are taken into account, they could be used as additional, integrated features in network-based data comparisons. Their tight relationship to edge existence probabilities can be exploited to predict missing or spurious edges in order to improve the network representation of the underlying biological system.", "keywords": [ "biomolecular networks", "co-expression networks", "edge relevance", "hierarchical stochastic block model", "missing and spurious edges", "module detection", "network-based omics analysis" ], "content": "Introduction\n\nHigh-throughput measurement techniques are advancing and become ever less expensive, enabling the screening of multiple biological data layers in single patients as almost standard clinical diagnostic tools. The wealth of the biological data can only be understood if treating the measured entities – gene promotors, mRNA, metabolites, proteins and their activity – not separately, but in their network context1. Thereby, one method to capture the interdependencies of the intracellular machinery relies on the hypothesis that strongly connected molecular entities are either co-activated or co-repressed, i.e. their measured abundances should be correlated2–4. Fully connected, weighted biomolecular networks can be established, in which each node corresponds to a molecular entity and is connected to each other node by an edge. The weight of the edge is the correlation between the measurements and is considered to represent how strongly the nodes are connected, interact or regulate each other.\n\nApproaching a network-level analysis of a biological system by correlation-based interactions has the advantage that it does not require a priori knowledge, and thus it is focused on the interaction profile for which evidence can be found in the measurements of the considered condition. However, this is a blessing and a curse: correlation-based networks suffer from the fact that the biological measurements are inherently noisy, even more so for small sample sizes as in rare diseases or in personalized medicine. This affects the values of the edge weights and the results of network reduction, and can be deleterious for subsequent analyses of the networks. In these cases, considering in addition alternative sources of edge relevance that are based on more global characteristics of the system might be beneficial and could make the network representations of the system and their analysis more robust.\n\nOne of the commonly performed analyses of correlation-based networks relies on the observation that biological networks exhibit a modular structure, in which tightly regulated modules are loosely connected to other modules. An important readout from correlation-based networks therefore is the biological and functional characterization of condition-specific modules, i.e. communities of co-regulated entities. A plethora of methods for module detection in networks has been proposed5–14. Also for the inference of edge relevance, or edge prediction, as being tightly linked to the problem of the detection of missing or spurious edges, numerous methods have been suggested15–23. In this work, we showcase a method that is able to derive, within a single framework, modules as well as scores of edge relevance: representing the biomolecular correlation-based networks as stochastic block models (SBMs)24.\n\nSBMs are the simplest form of generative network models for community structures and can also accommodate hierarchies25, which are key to convey robustness to biological function. Other generative model approaches relying on scale-freeness of the network architecture have been used for edge prediction in protein-protein interaction networks15, and the SBM has been used for representing other types of networks17,18,26,27, but not yet for biomolecular networks. In generative network model approaches, the network is described by a stringent mathematical framework based on statistical assumptions on network characteristics. For the case of the stochastic block model, this step delivers already the modular structure of the network as the nodes are assigned to blocks according to their connectivity properties to all other nodes. In contrast to other module detection methods, blocks in the SBM are not necessarily formed by tightly intra-connected entities but by entities which interact similarly with the nodes from all other blocks. Therefore, comparing SBM-derived modules between networks representing different (e.g. biomedical) conditions could be especially informative to shed light on regulatory changes. In a second step, the mathematical representation of the networks by SBMs is exploited to estimate edge probabilities. Specifically, it is assessed whether the existence of an edge in the network improves or reduces the fit of the network to the SBM. The resulting edge confidence scores are based on global network structure and can be used as alternative measure of edge relevance.\n\nHere we showcase the SBM-based analysis (overview in Figure 1A) for networks of three different molecular types, derived from transcriptomic, proteomic and metabolomics data of breast cancer tumours. We assessed which of the different versions of SBM fits each data layer best. Then, we investigated whether the SBM representation is able to capture functionally relevant structures in our biomolecular networks. In detail, we determined the agreement between the predicted blocks and independent biomolecular functional annotations from databases. Finally, we took advantage of the description of the networks as SBMs for the computation of an edge confidence score for each edge as measure of edge relevance. The edge confidence scores can be exploited to re-establish erroneously removed edges or to remove spurious edges, or they could serve by themselves for deriving disease-relevant differences when comparing groups of patients.\n\n(A) Pipeline for the approach. Given the matrix of measurements of mRNA expression, protein expression or metabolite abundance for a group of samples (I), we compute a fully-connected weighted network (II) of the molecular species for each data layer separately using a correlation-based approach. We reduce the networks by setting a threshold to the edge weights and binarize the networks (III). Each network is fitted to different types of stochastic block models in which the network nodes are partitioned into blocks (IV), the best fitting model is employed for deriving SBM-based edge confidence scores (V). (B) We established correlation-based networks using Spearman’s correlation within mRNA expression, protein expression and metabolite biomolecular data from a subgroup of cancer patients. Shown are histograms of the correlations obtained for all edges of the networks. (C) We only kept edges in the network that had a correlation which differed significantly from zero (sign. corr.). Based on this criterion, for different multiple testing correction methods (BH: Benjamini-Hochberg, BF: Bonferroni) and significance levels (0.05, 0.01), different degrees of reduction can be achieved. We chose for each data layer the most stringent threshold (highlighted in blue) that reduced the edge count to less than 107 edges (left) while keeping the percentage of nodes in the largest connected component high (right). (D) Histogram of the node degrees of the networks reduced by criterion on significance of correlation (red) or on scale-freeness (blue) on double log-scale. For scale-free networks, a linear relationship between log-frequency and log-node-degree is expected.\n\nAll code is freely available on https://gitlab.com/biomodlih/sbm-for-correlation-based-networks.\n\n\nMethods\n\nBreast cancer mRNA expression from RNAseq was obtained from the TCGA BRCA cohort via RTCGA28 downloading TCGA level 3 preprocessed BRCA files (search term: mRNAseq_Preprocess) on Nov 2, 2017. We used the normalized RSEM values. Protein data was obtained via the CPTAC homepage. We used the first replicate of samples measured in duplicates. We employed the unshared log ratio value for each sample to maximize reliability of protein identification. Clinical data for both TCGA and CPTAC data was retrieved and evaluated using the RTCGA package. Specifically, we used the following files for mRNA, protein and clinical data, respectively:\n\ngdac.broadinstitute.org_BRCA.mRNAseq_Preprocess.Level_3.2016012800.0.0/ BRCA.uncv2.mRNAseq_RSEM_normalized_log2.txt\n\nTCGA_Breast_BI_Proteome_CDAP_Protein_Report.r3/Protein_data/CDAP/TCGA_Breast_BI_Proteome.itraq.tsv\n\ngdac.broadinstitute.org_BRCA.Merge_Clinical.Level_1.2016012800.0.0/BRCA.clin.merged.txt.\n\nFor both mRNA and protein data, we only used samples from patients whose entry \"patient.breast_carcinoma_estrogen_receptor_status\" in their clinical data was \"negative\". In addition, we restricted our analysis to solid tumour samples, i.e. TCGA sample identifiers ending with 01. This gave rise to 237 samples for the mRNA data and 36 samples for the protein data. Metabolite data was used from the Excel file provided in 29 using the 67 samples containing ERn in their label.\n\nWe removed missing values in the mRNA data by replacing them by -10 to account for the fact that they arose from logarithmizing read counts of zero. In the protein data, we removed the 2195 genes which had more than 20% of missing values over the considered samples (i.e. more than 7 NAs among the 36 samples). The metabolite data did not contain missing values as imputation had been performed before29.\n\nWe used the correlation computation from the Hmisc package30 (function rcorr) to determine Spearman’s correlation of the measurements of each pair of entities (mRNA, protein or metabolite) over all samples. Only pairwise complete observations were employed. Unless stated otherwise, we neglected self-edges.\n\nThe Hmisc package30 was used to determine the p-value associated to the correlation (significance of the correlation being different from zero). We corrected for multiple testing using Bonferroni (mRNA expression) or Benjamini-Hochberg31 (protein, metabolite) correction. Correlations were considered significantly different from zero for corrected p-values lower than 0.05 (mRNA, metabolite) or 0.01 (protein). For the reduction by imposing a scale-free architecture of the reduced network, we employed the pick_hard_threshold function of the WGCNA package3 with the default requirement (0.85) on goodness of fit to a power-law degree distribution of the nodes.\n\nWe converted the adjacency matrices of the reduced networks to edge lists in csv format, added the disconnected nodes and fed the resulting networks as graphs into the Python graph-tool32 framework. Using an agglomerative multi-level Markov Chain Monte Carlo algorithm33, that module allows for determining a (potentially hierarchical) partition, b, of the network, G, which minimizes the description length, DL, of the SBM34:\n\n\n\nThereby, P(A\|C, D) denotes the probability of A given C and D, and λ captures all parameters of the model apart from the partition, such as the number of edges between blocks and degree distribution parameters for the degree-corrected SBM, in their planar or hierarchical version. Note that the above relationship holds only under a microcanonical formulation of the priors, i.e. hard constraints imposed on the values of the model parameters λ by the structure of the network G the SBM is fitted to (see 34), which enables considering only one value for the parameter λ for a given partition b and thus\n\n\n\nBecause the probability of the network itself, P(G), is constant for a fixed observed network, the description length is monotonously inversely related to the probability that partition b is responsible for the observed network G, P(b\|G). Therefore, finding the partition which minimizes the description length is equivalent to finding the partition with maximal posterior probability, P(b\|G). Furthermore, as can be seen in its second term in the first line, the description length DL contains a \"penalty\" term for the complexity of the SBM description. Thus, it can be used to distinguish which of the four examined SBM types (classical, hierarchical, degree-corrected, hierarchical+degree-corrected) is most suitable to describe the network G. Due to the large sizes of the networks, sampling over the posterior distribution is costly. Therefore, we decided to compare different SBM versions using only the estimated maximum of the posterior, i.e. at the partition b with the lowest minimal description length, in the expression of the posterior odds ratio34. Assuming that both SBM types that are compared, SBM1 and SBM2, are equally probable a priori, P(SBM1) = P(SBM2), we obtain for the posterior odds ratio Λ34:\n\n\n\nWe can consider the SBM type with the lowest minimal description length most likely; the distance of the posterior odds ratio to 1 determines how much more likely it is than the other SBM type.\n\nWe performed 500 runs with random initial partitions for each of the four SBM types and each of the six networks. The SBM type with lowest minimal description length was used for further analyses.\n\nWithin the mRNA and protein networks, biological annotation of the blocks and overrepresentation was performed using Reactome pathways and the package ReactomePA35. We restricted our analysis to Reactome pathways containing at least 10 and at most 500 annotated genes, all entities of the network were used as background. We employed Benjamini-Hochberg multiple-testing correction. Pathways were considered enriched for default settings (p-value < 0.05, q-value < 0.2), and only human pathways occurring in the file from the Reactome database, https://reactome.org/download/current/ReactomePathwaysRelation.txt (downloaded June 6, 2018), were used. This file was also employed as representation of the Reactome hierarchy to determine distances between Reactome pathways.\n\nFor the metabolite dataset, we mapped the pubchem IDs and metabolite names to KEGG compound IDs using the MBRole webserver version 236 and merged them semi-manually, thereby preferring metabolite names (in case of mismatch with pubchem record) and KEGG IDs with pathway annotation. We downloaded the human KEGG pathway annotation for the mapped metabolites from MBRole and used it as user-defined annotation for overrepresentation analysis with the enricher function from the package clusterProfiler37. We only considered pathways with a minimal size of 2 and otherwise the same settings as for mRNA and protein overrepresentation analysis.\n\nWe employed two distance measures of Reactome pathways considering the graph given by the Reactome familial hierarchy tree: (i) the distance in terms of number of steps necessary in the Reactome hierarchy graph to reach from one pathway to the other (distance 1), and (ii) the hierarchy level of the lowest common ancestor, or least common subsumer, for ontology graphs (distance 2). We incorporated an artificial top-Reactome pathway into the hierarchy to connect all pathways with each other and to have our distance measures well-defined. It lies at the 10th hierarchy level, so the largest distance between two pathways is 10 for distance 1 and 18 for distance 2 (nine steps to the highest level and nine back). For all comparisons, only blocks with at least one overrepresented pathway were considered. For comparing the distances of Reactome pathway annotations between blocks, i.e. to associate a distance to a pair of blocks, we median-averaged the distances over all combinations of Reactome pathways associated to the two blocks. For distances of pathway annotations within blocks, we median-averaged the distances between all possible combination of different Reactome pathways associated to the block. The trivial distances of zero for the distance of a Reactome pathway to itself were omitted, as well as blocks with only one enriched Reactome pathway for intra-block distances. Distances between blocks in the SBM hierarchy, as employed in Figure 3D, were defined analogously to the distance measure (i) based on the graph of the SBM hierarchy.\n\nAlternative clustering using the WGCNA package3 was performed following the WGCNA tutorial on clustering. We employed the full correlation matrices including self-edges. First, the correlation values were scaled to values between zero and 1 (by (1+corr)/2), and the soft threshold delivering scale free network topology was determined using the pickSoftThreshold.fromSimilarity function with default settings. Power estimates were 8, 16 and 12 for the mRNA, protein and metabolite network, respectively. We calculated the dissimilarity using TomSimilarity on the soft thresholded correlation matrix, used hclust with method \"average\" and cutreeDynamic with \"deepSplit\" parameter of 4 (mRNA) or 2 (else), \"pamRespectsDendro\" set to FALSE and \"minClusterSize\" of 3 (for mRNA, metabolite) or 4 (for proteins) to make the clustering most similar to the one obtained from the SBM.\n\nWe aim to derive edge confidence scores for each single edge in the network exploiting the representation of the network as SBM. Let us consider a fixed network given as graph G. If δG is a set of edges which do or do not occur in the network G, the probability that these edges belong to the observed network (for edges missing in G) or do not belong to the observed network (for edges from G), P(δG, G), can be written as27:\n\n\n\nfor b being partitions of the network G (please refer to Fit to SBM for further information on notation). The derivation assumes that the original network, G and the altered network with edges in δG added or removed, G + δG, has been generated by some SBM type (and all probabilities are conditional on that SBM type), and that the set δG has been chosen by some uniform distribution among all possible edges. The proportionality factor between the expressions depends on the network G and the number of edges in δG, and thus can be in particular neglected if only comparing edge confidences between single edges of a network. Because we aim to score all edges, and due to the sizes of the networks making computations slow, we refrained from sampling over the posterior distribution of the partitions and instead employed the single-point approximation for edge prediction proposed in 27:\n\n\n\nIt resorts to neglecting the summands for all partitions except for the one, b, which contributes most to the posterior distribution, b = maxbP(b\|G). In addition, the approximation relies on the assumption that the estimated optimal partition for the representation of G by the SBM is the same for G and its altered version with added or removed edge, G + δG, which is reasonable for our application case of single edge predictions, i.e. for δG being composed of a single putatively missing or spurious edge. The term P(b\|G) can be considered constant for a fixed G and SBM type, so we can shift it to the proportionality factor. Considering the microcanonical formulation of the SBM (see Fit to SBM), it becomes clear that the edge predictions for δG directly depend on the difference between the description length of the original network G with partition b, DLG,b, and the description length of the altered network G + δG with partition b, DLG+δG,b27:\n\n\n\nThe difference between the description lengths, DLG,b − DLG+δG,b*, was computed via the function get_edges_prob from graph-tool32. Note that as of time of writing, get_edges_prob does not work for weighted SBMs with real-normal edge covariate (see filed issue #452 at graph-tool.skewed.de). In addition, this function employs the natural (instead of the dual) logarithm and consequently, the obtained value has to be scaled by log2(e) to obtain the plain difference of description lengths.\n\nIn order to make clear that neither these edge predictions nor their dual (or natural) logarithm correspond to actual probabilities, we used the term edge confidence scores or simply edge scores throughout the manuscript.\n\n\nResults\n\nWe showcase the methods for data from a subgroup of breast cancer patients: those with estrogen receptor negative (ER-) tumours. ER status is predictive of patient outcome, with ER- leading to unfavourable prognoses, and its assessment is part of standard breast cancer patient screening38–40. The Cancer Genome Atlas (TCGA) initiative has provided data on the mRNA level (from RNAseq) for several hundred patient tissue samples; mass-spectrometry-based proteomic data (4plex iTRAQ) are available for a subgroup of 36 patients41. Metabolomics data have been measured by GC-TOF-MS in a different breast cancer cohort study for 67 samples29.\n\nWe treated each measured entity as node and established a correlation-based, weighted, biomolecular network for each single measurement layer. Therein, each pair of nodes is connected by an edge for which the weight is determined by Spearman’s correlation of the measurements of the nodes (over the samples), delivering values between -1 and 1. Thus, an edge that connects nodes with a correlation close to 1 or close to -1 represents strong positive regulation and strong negative regulation, respectively; edges connecting nodes with a correlation close to zero represent weak or absent regulation. We employed Spearman’s correlation because it captures also non-linear relationships between measurements, and it is robust to outliers and any monotonous transformation (e.g. logarithmization). We dealt with missing values in the data by replacing them by small values (NAs due to log transformation of zero counts in the mRNA data) or removing entities with >20% missing values for all samples (for the protein data). Subsequently, we computed the correlation only considering pairwise complete observations. The distributions of the computed correlation values for the three data types are shown in Figure 1B.\n\nThe resulting biomolecular networks capture the relationships of the intracellular machinery, and their analyses deliver important insights on altered regulations in disease states2,3. However, because these networks are fully connected, i.e., every entity is connected to each other entity within a layer, the networks become very large and their analyses difficult. A common approach is to reduce the networks, either by selecting a subset of entities as nodes prior to network establishment, mainly by using criteria on abundance, or by using the assumption that weak regulations are less important for the biological network and can be omitted without impairing the represented function of the network. We decided on the latter approach in order not to bias our choice of considered molecular entities.\n\nWe used two different techniques of network reduction by thresholding: In the first, we only kept edges for which the correlation was significantly different from zero (\"sign. corr.\"), i.e., the regulation being sufficiently strong. We adapted the significance thresholds for the different molecular data layers to their respective sizes. Larger networks were subjected to more stringent cut-offs in order to achieve an appropriate degree of reduction (Figure 1C, employed thresholds marked in blue). In the second reduction method, we removed weak links until the reduced network met a criterion of scale-freeness (3, \"scale-free\"). Scale-freeness is considered a key property of self-organized networks and thus also of biological molecular networks1. In scale-free network formation, highly connected nodes tend to attract more connections than lowly connected nodes leading to a degree distribution following a power-law; the degree distributions of the reduced networks are shown in Figure 1D. Both reduction techniques are hard-thresholding techniques meaning that edges are removed from the networks. The resulting six reduced networks, two for each data layer, were used in a binary form, i.e., weight information was discarded after reduction. Some characteristics of the original and reduced networks are shown in Table 1; the node count of the reduced networks coincides with the number of entities considered after missing value removal. In the following sections, we describe how to analyze these different homogeneous cancer networks by fitting them to SBMs.\n\nNote that genes/proteins/metabolites were removed if having >20% NA values (which was the case only for proteins), all other entities were kept as nodes in the reduced network even if their degree was zero (i.e. having no edge) after network reduction.\n\nBiological networks are known to be modular and hierarchical. Different molecular entities, such as genes, mRNAs, and proteins, are interconnected and form different modules to fulfill a specific function. Modularity can convey more robustness to the overall system, e.g. by preventing perturbations in single modules to spread fast and to cause erroneous behavior in other modules and thus functions. Hierarchies capture two characteristics of biological systems: (i) the ordered combination of functions, i.e., multiple simple functions resulting in more complex behavior or responses; and (ii) the inherent levels of complex organization of life, from single molecules to cell organelles, cells, tissue, organs and whole organisms.\n\nThe stochastic block model (SBM) is the simplest form of a generative network model based on communities, i.e., group structures of the nodes. Thereby, nodes are assigned to blocks according to their connectivity properties (Figure 2A left); the block associations of two nodes fully determine the probability of an edge between them. A shortcoming of the classical SBM for representations of real networks is that nodes within one block need to have similar degrees. The degree-corrected version of the SBM42 accounts for that and enables different degrees for nodes within a block. Another extension of the SBM is its hierarchical version, in which the blocks are further partitioned into blocks of higher levels25 (Figure 2A right). This model is especially suitable to represent large networks with many nodes. Fitting biological networks to hierarchical (also called nested) SBMs is therefore most appropriate.\n\nWe assessed which of the four following types of stochastic block models could best represent the biological networks: the classical SBM, the degree-corrected SBM, the hierarchical SBM or the degree-corrected and hierarchical SBM (Figure 2B). We used the Python module graph-tool32 to fit SBMs to the networks, i.e., to find which partition (basal and/or hierarchically ordered) describes the network best as SBM.\n\n(A) In a stochastic block model (SBM) representation, the nodes of a network are partitioned into blocks according to their similarity in connectivity. The hierarchical version of the SBM (right) imposes in addition hierarchical partitions onto the blocks. (B) Six biomolecular networks derived from transcriptomic, proteomic or metabolomics data of breast tumours were fitted to four different types of SBM: the classical SBM (SBM, light blue), the degree-corrected SBM (dcSBM, light green), the hierarchical SBM (hSBM, dark blue) and the degree-corrected hierarchical SBM (dchSBM, dark green). We performed fits for 500 initial partitions for each network. A fit consists of altering the partitions underlying the SBM such as to minimize the description length. The smaller the description length the better the fit. Hierarchical SBMs outperform non-hierarchical SBMs, degree correction is required for the mRNA and protein networks. (C) Graphical representations of the best fitting hierarchical stochastic block models with degree correction (dchSBM) or without (hSBM) for the networks reduced by significance of correlation (mRNAs, metabolites) or by scale-freeness (protein). The lowest layer (genes, proteins) is truncated in the mRNA and protein networks, colored lines denote edges between blocks of the first level (for mRNA, protein network) or between metabolites (level 0, metabolite network). Edges of blocks or metabolites which belong to the same block in the level above have the same colour. The higher-order hierarchical structure is shown in black.\n\n(A) Percentage of blocks with at least one overrepresented Reactome (KEGG) pathway for the best fitting SBM (see Figure 2B) for the mRNA and protein (metabolite) networks, reduced by condition on scale-freeness (scale-free) or on significance of correlation (sign. corr.), for each hierarchical level (bars). Black crosses denote the percentages of blocks with at least one overrepresented Reactome (KEGG) pathway for three SBMs each with exactly the same structures but randomly shuffled mRNAs or proteins (metabolites). (B) For the lowest hierarchy level clustering of each of the four mRNA and protein SBMs, we calculated the average distances between every pair of Reactome pathways between blocks and those within blocks, for two different distance measures based on the Reactome hierarchy. The lower the distances the more similar are the pathways. For both distance measures, the pathways associated to one single block (within) are significantly more similar than those associated to different blocks (Welch’s t-test p-value < 0.01) suggesting that biological functions are consistent within blocks and distinct between blocks. (C) Distances of Reactome pathways (as in B) between blocks (or within blocks, for distance of blocks in SBM being zero) versus the distance of the blocks in the SBM. We do not find evidence that the Reactome hierarchy is reflected in the SBM hierarchy. (D) Distribution of predicted module (block) sizes for the six networks using SBM (lowest hierarchical level only) or the three networks using network generation and module detection from WGCNA3. (E) Percentage of modules detected by the WGCNA approach with at least one enriched Reactome or KEGG pathway. Black crosses denote the results for three similar clusterings with randomly shuffled mRNAs, proteins, or metabolites.\n\nNote that we also examined the performance of weighted stochastic block models for our purpose of edge prediction as they have been successfully employed before for non-biomolecular networks18,27,43. However, the characteristics of the optimal weighted SBM (number of blocks) were severely impacted by prior assumptions on the edge weight distributions, and the derived edge confidence scores did not coincide well with evidence on edge relevance given by the correlations of the edges for fully-connected weighted networks (see Figure S144). Taking in addition the increased computational effort for fitting the weighted SBM compared to the binary version into account, we restricted our further analyses to non-weighted SBMs.\n\nThe model fit is performed following the rationale of Occam’s razor: The simplest model describing the data is the best. Thus, we searched for the partition that minimized the description length, i.e., the amount of information necessary to describe the network as an SBM. Additional information required to capture degree-correction and/or hierarchies compared to the classical SBM is thereby taken into account. Consequently, it can be directly concluded which of the four SBM model types is most appropriate for a certain network: The one with the lowest minimal description length. The optimization of the description length runs via an agglomerative Markov Chain Monte Carlo algorithm33. It is non-deterministic and multiple initiations of the underlying partition of the SBM are required to obtain globally instead of locally optimal partitions. We performed optimizations for 500 initial partitions for each network and SBM type.\n\nFor all four mRNA and protein networks, the classical SBM delivered the worst fit and the degree-corrected, hierarchical SBM fitted best (Figure 2B left and middle). Degree correction did not prove necessary to describe the metabolite networks, for which the hierarchical SBMs fitted slightly better than the classical SBM (Figure 2B right). A graphical representation of the best fitting SBMs in circular layout, showing the blocks from the lowest layer (for mRNA, protein) or the metabolites (metabolite network) and their connections in color, and the hierarchical structure on top in black, is given in Figure 2C.\n\nWe wanted to assess how well the biological content of the biomolecular networks is captured in the stochastic block models, i.e., if the clustering of the nodes into the blocks is biologically meaningful. To that purpose, we estimated whether the blocks show common biological function based on Reactome or KEGG pathways. In particular, we performed overrepresentation analyses of Reactome pathways for the blocks at each level of the SBMs of the four protein and mRNA networks, and overrepresentation of KEGG pathways for the blocks predicted for the two metabolite networks. We find that a high percentage of blocks in each level has at least one associated Reactome or KEGG pathway, i.e., the genes of the pathway are overrepresented within the block, they occur more frequently than expected by chance (Figure 3A bars). Except for the highest hierarchy levels that consist only of a few blocks, this percentage is decisively higher than for a random clustering of exactly the same structure (results for 3 random clusterings shown as black crosses in Figure 3A).\n\nMany blocks, however, have not only one but multiple Reactome pathways assigned. If blocks are biologically meaningful, we would expect to observe similar Reactome pathways within blocks but less similar pathways when comparing the pathways of different blocks.\n\nThe dissimilarity of Reactome pathways is naturally represented by their distance in the Reactome familial hierarchy structure; the more distant pathways are, the less connected are their biological functions. We used two closely related measures to assess it: (i) the distance of two pathways is the length of the shortest path in the hierarchy tree from one pathway to the other, and (ii) the higher the hierarchy level of the lowest common ancestor (least common subsumer) of two pathways, the more distant they are. Please note that we restricted this analysis to the mRNA and protein networks because the KEGG pathways employed for metabolite data are only assorted into a very shallow hierarchy.\n\nWe compared the distances of Reactome pathways using the average of the two distance measures over pairs of Reactome pathways associated with one block (within blocks) and over pairs of Reactome pathways associated with two different blocks (between blocks) for the lowest hierarchy level blocks, as shown in Figure 3B. Thereby, for all four networks and both measures, we observe significantly smaller distances of Reactome pathways within blocks than between blocks (Welch’s test p < 0.01). This suggests that biological function is coherent within and distinct between blocks, thus further enhancing the notion that SBMs represent well the biological function of the networks.\n\nWe also examined the relationships of the hierarchies within the SBMs to that of the Reactome pathways. In analogy to the distance of Reactome pathways, we defined the distance of blocks within a hierarchical SBM by counting the number of steps necessary to reach a block from the other (via their lowest common higher-level block). We compared the distance of each pair of blocks within the SBM on the lowest level to the average of the pairwise distances of their associated Reactome pathways (Figure 3C). Therein, we found only a weak positive correlation for Reactome pathway distance measure (i), that is even further reduced if neglecting intra-block coherence, i.e., neglecting blocks with distance zero. We concluded that the hierarchy within the SBM does not strongly coincide with the hierarchy within the Reactome pathways.\n\nIn order to assess how the clustering by SBM relates to established clustering techniques in correlation-based networks, we also performed module detection by using the WGCNA package3. Note that for this approach, no model reduction is necessary, which means that we obtain only one result for each data layer. After soft thresholding to enforce scale-free network architecture, we employed WGCNA module detection to obtain comparable numbers of modules as for the SBM-based approach: 333 (mRNA), 109 (protein) and 12 (metabolite) for WGCNA, which are similar to the 438, 113 and 15 blocks detected by SBM for the corresponding scale free networks. Overall, the WGCNA clusterings show a larger diversity of module sizes than those obtained for the SBM approach (Figure 3D). For all three data layers, a higher percentage of WGCNA-derived modules have a biological annotation compared to the blocks from the SBM (compare Figure 3E to 3A). However, for the mRNA layer, many of the blocks also have an enriched pathway annotation for randomly shuffled gene names which hints on reduced significance of the WGCNA results and better performance for the modules detected by fitting to SBM. For protein and metabolites, in terms of detection of biologically annotated modules, the WGCNA approach seems to provide slightly better results than the SBM approach.\n\nThe observed differences could stem from conceptual differences in the approaches: Instead of detecting clusters of entities with highly positively correlated abundances only as in WGCNA, nodes from the same block are characterized by common connectivity characteristics in the SBM. Clearly, as proteins interact and bind directly for complex formation or regulation, and metabolites are interconverted into one another, entities which act together tend to have similar abundances and thus the modularization by WGCNA shows good results. For the detection of modules for mRNAs whose interaction can be considered less direct, assorting entities with similar connectivity patterns as in the SBM-derived modules is beneficial. In addition, the WGCNA framework, as most other module detection methods, cannot aid in assessing edge relevance - which is enabled by the fit of the networks to SBMs.\n\nThe description of the biomolecular networks as SBMs were exploited to determine a confidence score for each existing or non-existing edge of the network. This score would capture how well the existence (or non-existence) of the edge fits to the network’s description by the SBM: Erroneously kept or removed edges lead to a worse fit of the SBM to the network, and removing or reinstalling these edges lead to fit improvement. Indeed, under certain assumptions (see Methods), the derived edge confidence score is proportional to the actual absolute probability that the edge belongs to the network, and thus signifies relative edge relevance. Therefore, the confidence score could be used to predict whether an edge is spurious (or missing)17,18,27. A high missing edge confidence score suggests that the edge in question is missing and should be restored, it has a high relevance; a high spurious edge confidence score suggests that the edge in question is spurious and should be removed from the network, it has a low relevance. The SBM-based edge confidence scores rely on global network connectivity characteristics and complement the correlation-based weights of the edges which stem from the local, measured characteristics of their nodes.\n\nFor each putatively missing and spurious edge, we used the Python module graph-tool32 to compute its edge confidence score (Figure 4). Thereby, we took advantage of the following fact: Entities of degree zero, i.e., nodes that have no connection to any other node in the network after reduction, are indistinguishable to the SBM. Consequently, also all putatively missing edges connecting any of these nodes to a specific second node are only distinguishable by this second node, and thus carry the same missing edge confidence score. This reduces the number of scores we need to compute considerably, depending on the degree of reduction (see counts of entities of degree zero in Table 1). We display the scores for the different types of missing edges in different histograms (Figure 4A, B middle and bottom). For very big networks with a low degree of reduction (the mRNA and protein networks reduced by significance of correlation), it is still computationally not feasible to compute the scores for all missing edges. We therefore resorted to computing it for as many missing edges as we have existing edges in the network (i.e. approx. 2.4×106 for the protein network, 4.6×106 for the mRNA network, see Table 1), and chose those with highest absolute edge weights (Figure 4B left and middle).\n\n(A) Histograms of edge confidence scores for the best fitting SBM of the three networks reduced by criterion on scale-freeness for all existing, i.e., putatively spurious edges (spurious edge score, top), for the putatively missing edges between nodes with degree > 0 (missing edge score, middle), for the putatively missing edges connecting a node of degree 0 to a node with a larger degree (missing edge (disc) score, bottom – only one edge for each node with degree > 0 is shown). (B) Edge confidence scores as described in (A) for the networks reduced by criterion on significance of correlation. For the mRNA and protein network, missing edge scores (middle) were only computed for the edges with largest absolute correlations of the node measurements.\n\nRecall that the edge confidence scores are relative, i.e., they serve for comparing relevance between edges only. In addition, computation of the scores relies on the assumption that the partition of the originally fitted SBM is correct for the network. Different edge confidence scores might be obtained for SBMs with different partitions but with similarly good fit to the network. We neglect this complication for the sake of computational efficiency. Still, we have to keep both facts in mind for the interpretation and usage of the edge confidence scores. For example, an evident threshold for declaring an edge as relevant (\"missing\") or not relevant (\"spurious\") would be to have the respective edge confidence score larger than zero, as this indicates an improvement in fit quality if adding or removing the edge, respectively. However, we observe an imbalance of our computed scores which is inherent to the approach: Because having less edges reduces the complexity of the underlying network, removing edges preferentially reduces also the amount of information required to describe the SBM, i.e., its description length turns smaller, its goodness of fit improves. Therefore, the edge confidence score distributions of missing edges are shifted to the left - the great majority of edges would be predicted as not missing for the threshold of zero, they are not relevant and should be left out (Figure 4A, B, 2nd line). Conversely, the spurious edge confidence score distributions are shifted to the right - the great majority of existing edges would be predicted as spurious, they are not relevant and should be removed for an edge score threshold of zero (Figure 4A, B, top). Both measures point to making the networks smaller.\n\nIn order to determine whether the edge confidence scores are overall a reasonable assessment of edge relevance, we compared the predicted scores to the edge weights of the edges as derived directly from Spearman’s correlation of the measurements of the nodes. It is important to note that these edge weights (correlations) were used exclusively for the reduction of the correlation-based networks. The edge correlations were at no point provided to the SBM, neither during the fitting to the SBM (except for the weighted version and only for Figure S1, Figure S244) nor during the computation of the edge scores. Thus, they are close to an independent validation of the edge relevance, edges with large correlation being more relevant to the system encoded by the network than edges with a correlation close to zero.\n\nIndeed, for all six networks and SBMs, we find an overall positive correlation between the (absolute) edge correlations and the missing edge confidence scores: Edges with high correlation are preferentially predicted as missing, i.e., relevant to the network, also in terms of edge confidence score (Figure 5). Similarly, we find an overall negative correlation between (absolute) edge correlations and spurious edge confidence scores: Edges with high correlation are preferentially not predicted as spurious, i.e., they are predicted as relevant to the network, also in terms of edge confidence score. Consequently, the comparison to the edge correlation suggests that SBM-derived edge confidence scores could be used as additional information for assessing the relevance of edges for multiple omics correlation-based networks.\n\nRelationship between edge confidence scores and edge correlations (Spearman’s correlation of the measurements at their nodes) for putatively spurious (top) or putatively missing edges between nodes of degree > 0 (bottom) for the networks reduced by criterion on scale-freeness (A) or significance of correlation (B). Edge confidence scores correlate well to edge correlations, with edges with high SBM-predicted confidence in spuriousness – edges predicted as irrelevant – having lower correlations, edges with high SBM-predicted confidence in missingness – edges predicted as relevant – having higher correlations. Please note that for the protein network reduced by significance of correlation, due to missing values in the dataset, a fixed significance threshold leads to different correlation thresholds and thus the correlation value boundary is blurred (B, middle).\n\n\nDiscussion\n\nUsing example cases of correlation-based transcriptomic, proteomic and metabolomics networks from breast cancer tumour samples, here we show how stochastic block models can be employed for the analysis of biomolecular networks. The networks can be best represented by the hierarchical version of the stochastic block models. This gives rise to biologically meaningful separation of the biomolecules into many functionally relevant blocks. In addition, the SBM framework enables the computation of edge confidence scores that can be used to predict missing or spurious links.\n\nThe representation of the networks by SBMs poses a challenge: The model fit and the derivation of edge confidence scores can be computationally very demanding, especially for networks with many edges. This was the case here for the mRNA networks and the protein network reduced by significance of correlation. Therefore, the approach of network analysis by SBM seems most suitable for smaller and/or sparser networks. On the other hand, it delivers two opportunities. First, modules derived from blocks in SBMs are not only defined as clusters of tight relationships, i.e. co-expression clusters, as obtained with other clustering approaches3,9,14. In an SBM, nodes from the same block are characterized by common connectivity characteristics, i.e., common interaction profiles with nodes from the same and from other blocks. Consequently, comparing SBM-derived modules between conditions naturally points towards detecting altered regulations. This has not been investigated here, the biological role and disease-relevance of the blocks in our example case of ER- breast cancer still need to be assessed.\n\nSecond, the derived edge predictions can reproduce the interaction strengths as estimated from measurements. SBM-based edge confidences tend to score missing edges higher that have high correlation, i.e., that would be considered a strong regulation, an important edge. Analogously, existing edges in the network with low correlation tend to be predicted as spurious with high confidence, i.e., as dispensable. Keep in mind that the correlations corresponding to the edges were not provided at any point to the SBM construction. Thus, the SBM approach proves very strong in delivering relevant edge predictions. A further biological validation of the predicted edge confidence scores, for example by comparison to interaction databases, would be an interesting next step.\n\nStill, a question remains: How should the edge confidence scores be translated into edge predictions to alter the network, i.e. to actually remove or re-install edges? There is a natural threshold for declaring an edge as missing or spurious, namely if the respective confidence score would be larger than zero. However, due to the minimal description length approach for SBM-based edge relevance assessment, the confidence scores are shifted towards reducing the networks as much as possible (Figure 4), such that this natural threshold for deriving the prediction of actual missingness or spuriousness from the confidence scores is not valid. Further examinations on other possible thresholds are required. However, for relative comparisons of relevance between edges in a network the SBM-based scores are suitable.\n\nAdditional directions of SBM-based analysis of biomolecular networks remain to be explored.\n\n(i) For edge relevance assessment in other network types, it has been proposed that edge predictions with SBMs are more reliable if resorting to an ensemble of good fits instead of using the best fit only27. Due to the sizes of the employed correlation-based biomolecular networks, this approach is computationally not of practical relevance here, but it would be an interesting point to assess in the future.\n\n(ii) We considered the weighted version of the SBM that could be an interesting option for the analysis of biomolecular networks because it enables representing fully-connected weighted networks as SBM without prior reduction. It could prove exciting especially for smaller networks, e.g. primary metabolites, or in more targeted data analysis approaches. However, in our case, we found hints that the weighted version of the SBM might not be appropriate for the task of edge prediction from SBM-based confidence scores for fully-connected networks (see Figure S144). For reduced networks, the results for weighted SBMs seem more promising (see Figure S244). A final assessment on the usefulness of the weighted version of the SBMs is still pending, as long as nothing is known about interaction strength distributions between and within modules: The fitted SBM strongly depends on the prior assumptions chosen for these distributions. An extensive analysis of different choices would be required, which is beyond the scope of this study.\n\nWe propose to employ SBM-based modules and edge confidence scores as additional pieces of information to make the results of follow-up analyses more robust that rely on relationships between nodes, i.e., on edges and their strength. Edges and their strength play a key role in interpreting biomolecular effects, e.g. of mutations or drugs. If a drug “activates a protein”, this typically means alteration of the protein’s interaction strengths with other proteins or the DNA. The “mutation of a gene” may severely alter the binding properties of the corresponding translated protein to interaction partners. In sum, interactions and their strengths are at the heart of biologically relevant alterations in biomolecular networks, and characterizing the former is required in order to understand the effects of the latter. The SBM framework enables assessing the edge relevance or interaction strength on the basis of consistent, global network characteristics, instead of on the basis of correlation of measurements. Especially for personalized analyses which generally rely on a characterization by only few error-prone measurements of each molecule, this will be crucial to derive more reliable predictions.\n\n\nData availability\n\nIn this study, data from TCGA and CPTAC was used. Proteomics data stem from 41, metabolomics data stem from 29. The metabolomics raw data can only be obtained upon accessing the cited article29; processed data (Spearman’s correlations and associated p-values) can be found in the gitlab repository below.\n\nCode used to perform the analyses together with a detailed work-flow documentation: https://gitlab.com/biomodlih/sbm-for-correlation-based-networks\n\nArchived code as at time of publication: http://doi.org/10.5281/zenodo.262041944\n\nLicense: GNU GPLv3 license.\n\nZenodo: Analysis of correlation-based biomolecular networks from different omics data by fitting stochastic block models. http://doi.org/10.5281/zenodo.262041944. This project contains the following extended data files:\n\nFigure S1: The weighted SBM seems not appropriate for edge prediction from edge confidence scores for fully connected networks\n\nFigure S2: Edge predictions for a reduced weighted network with planar or hierarchical weighted SBMs\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).", "appendix": "Grant information\n\nThis research was funded by Luxembourg National Research Fund, FNR, SINGALUN project. KB acknowledges funding by an Add-on Fellowship for Interdisciplinary Life Sciences of the Joachim Herz Stiftung.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nBarabási AL, Gulbahce N, Loscalzo J: Network medicine: a network-based approach to human disease. Nat Rev Genet. 2011; 12(1): 56–68. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nvan Dam S, Võsa U, van der Graaf A, et al.: Gene co-expression analysis for functional classification and gene-disease predictions. Brief Bioinform. 2018; 19(4): 575–592. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLangfelder P, Horvath S: WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008; 9: 559. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nToubiana D, Xue W, Zhang N, et al.: Correlation-Based Network Analysis of Metabolite and Enzyme Profiles Reveals a Role of Citrate Biosynthesis in Modulating N and C Metabolism in Zea mays. Front Plant Sci. 2016; 7: 1022. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPellegrini M: Community Detection in Biological Networks. Encyclopedia of Bioinformatics and Computational Biology. Elsevier, 2019; 1: 978–987. Publisher Full Text\n\nLangfelder P, Horvath S: Fast R Functions for Robust Correlations and Hierarchical Clustering. J Stat Softw. 2012; 46(11): pii: i11. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJeub LGS, Sporns O, Fortunato S: Multiresolution Consensus Clustering in Networks. Sci Rep. 2018; 8(1): 3259. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLancichinetti A, Fortunato S: Consensus clustering in complex networks. Sci Rep. 2012; 2: 336. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSu Y, Wang B, Zhang X: A seed-expanding method based on random walks for community detection in networks with ambiguous community structures. Sci Rep. 2017; 7: 41830. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRosvall M, Bergstrom CT: An information-theoretic framework for resolving community structure in complex networks. Proc Natl Acad Sci U S A. 2007; 104(18): 7327–31. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nReichardt J, Bornholdt S: Statistical mechanics of community detection. Phys Rev E Stat Nonlin Soft Matter Phys. 2006; 74(1 Pt 2): 016110. PubMed Abstract \| Publisher Full Text\n\nNewman ME, Girvan M: Finding and evaluating community structure in networks. Phys Rev E Stat Nonlin Soft Matter Phys. 2004; 69(2 Pt 2): 026113. PubMed Abstract \| Publisher Full Text\n\nEnright AJ, Van Dongen S, Ouzounis CA: An efficient algorithm for large-scale detection of protein families. Nucleic Acids Res. 2002; 30(7): 1575–84. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBlondel VD, Guillaume J, Lambiotte R, et al.: Fast unfolding of communities in large networks. J Stat Mech Theory Exp. 2008; 2008(10): P10008. Publisher Full Text\n\nZhu Y, Zhang XF, Dai DQ, et al.: Identifying spurious interactions and predicting missing interactions in the protein-protein interaction networks via a generative network model. IEEE/ACM Trans Comput Biol Bioinform. 2013; 10(1): 219–25. PubMed Abstract \| Publisher Full Text\n\nWang H, Zhang F, Min H, et al.: SHINE: Signed heterogeneous information network embedding for sentiment link prediction. Proceedings of the Eleventh ACM International Conference on Web Search and Data Mining. 2018; 592–600. Publisher Full Text\n\nGuimera R, Sales-Pardo M: Missing and spurious interactions and the reconstruction of complex networks. Proc Natl Acad Sci U S A. 2009; 106(52): 22073–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAicher C, Jacobs AZ, Clauset A: Learning latent block structure in weighted networks. J Complex Netw. 2015; 3(2): 221–248. Publisher Full Text\n\nWilliamson SA: Nonparametric network models for link prediction. J Mach Learn Res. 2016; 17. Reference Source\n\nZhu B, Xia Y, Zhang XJ: Weight prediction in complex networks based on neighbor set. Sci Rep. 2016; 6: 38080. PubMed Abstract \| Publisher Full Text\n\nNavlakha S, Gitter A, Bar-Joseph Z: A network-based approach for predicting missing pathway interactions. PLoS Comput Biol. 2012; 8(8): e1002640. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShakibian H, Moghadam Charkari N: Mutual information model for link prediction in heterogeneous complex networks. Sci Rep. 2017; 7: 44981. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPan L, Zhou T, Lü L, et al.: Predicting missing links and identifying spurious links via likelihood analysis. Sci Rep. 2016; 6: 22955. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHolland PW, Laskey KB, Leinhardt S: Stochastic blockmodels: First steps. Soc Networks. 1983; 5(2): 109–137. Publisher Full Text\n\nPeixoto TP: Hierarchical block structures and high-resolution model selection in large networks. Phys Rev X. 2014; 4(1): 011047. Publisher Full Text\n\nZhang X, Wang XJ, Zhao CL, et al.: Degree-corrected stochastic block models and reliability in networks. Physica A-Statistical Mechanics and Its Applications. 2014; 393: 553–559. Publisher Full Text\n\nVallès-Català T, Peixoto TP, Sales-Pardo M, et al.: Consistencies and inconsistencies between model selection and link prediction in networks. Phys Rev E. 2018; 97(6–1): 062316. PubMed Abstract \| Publisher Full Text\n\nKosinski M, Biecek P: RTCGA: The cancer genome atlas data integration. 2016. Reference Source\n\nBudczies J, Brockmöller SF, Müller BM, et al.: Comparative metabolomics of estrogen receptor positive and estrogen receptor negative breast cancer: alterations in glutamine and beta-alanine metabolism. J Proteomics. 2013; 94: 279–88. PubMed Abstract \| Publisher Full Text\n\nHarrell FE Jr and with contributions from Charles Dupont and many others: Hmisc: Harrell miscellaneous. 2018. Reference Source\n\nBenjamini Y, Hochberg Y: Controlling the false discovery rate a practical and powerful approach to multiple testing. J R Statist Soc B. 1995; 57(1): 289–300. Reference Source\n\nPeixoto TP: The graph-tool python library. figshare. 2014. Publisher Full Text\n\nPeixoto TP: Efficient Monte Carlo and greedy heuristic for the inference of stochastic block models. Phys Rev E Stat Nonlin Soft Matter Phys. 2014; 89(1): 012804. PubMed Abstract \| Publisher Full Text\n\nPeixoto TP: Bayesian stochastic blockmodeling. eprint arXiv:1705.10225. 2017; arXiv:1705.10225. Reference Source\n\nYu G, He QY: ReactomePA: an R/Bioconductor package for reactome pathway analysis and visualization. Mol Biosyst. 2016; 12(2): 477–9. PubMed Abstract \| Publisher Full Text\n\nLópez-Ibáñez J, Pazos F, Chagoyen M: MBROLE 2.0-functional enrichment of chemical compounds. Nucleic Acids Res. 2016; 44(W1): W201–W204. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYu GC, Wang LG, Han YY, et al.: clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012; 16(5): 284–287. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHarvey JM, Clark GM, Osborne CK, et al.: Estrogen receptor status by immunohistochemistry is superior to the ligand-binding assay for predicting response to adjuvant endocrine therapy in breast cancer. J Clin Oncol. 1999; 17(5): 1474–81. PubMed Abstract \| Publisher Full Text\n\nSamaan NA, Buzdar AU, Aldinger KA, et al.: Estrogen receptor: a prognostic factor in breast cancer. Cancer. 1981; 47(3): 554–60. PubMed Abstract\n\nClinical practice guidelines for the use of tumor markers in breast and colorectal cancer. Adopted on May 17, 1996 by the American Society of Clinical Oncology. J Clin Oncol. 1996; 14(10): 2843–2877. PubMed Abstract \| Publisher Full Text\n\nMertins P, Mani DR, Ruggles KV, et al.: Proteogenomics connects somatic mutations to signalling in breast cancer. Nature. 2016; 534(7605): 55–62. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKarrer B, Newman MEJ: Stochastic blockmodels and community structure in networks. Phys Rev E. 2011; 83(1): 016107. Publisher Full Text\n\nPeixoto TP: Nonparametric weighted stochastic block models. Phys Rev E. 2018; 97(1–1): 012306. PubMed Abstract \| Publisher Full Text\n\nBaum K, Rajapakse JC, Azuaje F: Analysis of correlation-based biomolecular networks from different omics data by fitting stochastic block models (version v2) [dataset]. Zenodo. 2019. http://www.doi.org/10.5281/zenodo.2620419" }	[ { "id": "47226", "date": "26 Apr 2019", "name": "Lei Xie", "expertise": [ "Reviewer Expertise bioinformatics", "systems biology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper by Baum et al describes a method to construct biological networks from omics data using stochastic block models (SBMs). The application of SBM on mRNA, protein and metabolic data gives a new way of deriving information from correlation based biological networks. The proposed method could be a useful addition to network biology. To strengthen the manuscript, I would suggest the following points to be addressed:\nIn the over representation analysis, while the statistics on distance of Reactome/KEGG terms shows the block generated here is significantly better than random. Is it possible to provide a such comparison between the clustering result by SBM and WGCNA? In the network reduction step, a linear relationship is expected between log-frequency and log-node-degree. While the scale-free fit index (R2) will be > 0.85 with WGCNA default requirement, how different threshold affect that linearity is not clear from Fig 1D. Also, how well can the network be reduced by significance of correlation fit to a scale free network? To demonstrate that SBM can provide biological insight, more detailed analysis on the benchmark data sets could be useful. For example, what are unique and common pathways for different breast cancer subtypes, how well are these findings consistent with existing knowledge? Figure 4 shows SBM based confidence score can be used to predict the existence of an edge. However, it is not clearly stated how the edges are determined as putatively missing or spurious. Also, the description for figure 4A is hard to understand.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4831", "date": "27 Aug 2019", "name": "Katharina Baum", "role": "Author Response", "response": "We thank Prof. Lei Xie and Yue Qiu for their comments that helped improving the accessibility of the presented contents and strengthened the manuscript. In the following, we answer to the specific points they raised and how we addressed them in version 2. >The paper by Baum et al describes a method to construct biological networks from omics data using stochastic block models (SBMs). The application of SBM on mRNA, protein and metabolic data gives a new way of deriving information from correlation based biological networks. The proposed method could be a useful addition to network biology. To strengthen the manuscript, I would suggest the following points to be addressed: >In the over representation analysis, while the statistics on distance of Reactome/KEGG terms shows the block generated here is significantly better than random. Is it possible to provide a such comparison between the clustering result by SBM and WGCNA? Response: We thank the reviewers for their appreciation of our work. We now also provide the computation of the SBM distances for the mRNA and protein WGCNA clustering results (see new Fig. 3F and new Fig. S3C). They show also a bigger distance between clusters than within clusters. We added the resulting p-values to the comparison of the distances for both SBM (Fig. 3B, D) and WGCNA (Fig. 3F, S3C) approach. >In the network reduction step, a linear relationship is expected between log-frequency and log-node-degree. While the scale-free fit index (R2) will be > 0.85 with WGCNA default requirement, how different threshold affect that linearity is not clear from Fig 1D. Also, how well can the network be reduced by significance of correlation fit to a scale free network? Response: Fig. 1D (now 1E) has been included for illustrative purposes only to show some further characteristics (in this case: degree distribution) apart from those in Table 1 of the six networks employed for the SBM-based analysis. It can be used to appreciate the fact that the networks reduced by requirement on significance of correlation are not optimized for it but still may exhibit a close to linear relationship between log-frequency and log-node-degree. For clarification, we added the scale-free fit indices obtained for each network into old Fig. 1D - now Fig. 1E. We also extended the description of the scale-freeness fitting procedure and approach by using the WGCNA package (see revised Methods section). Please note that less strict correlation thresholds than the employed here lead to networks being further away from a network with scale-free characteristic (this is inherent to the WCGNA fitting approach as the least stringent threshold is determined leading to a scale-free fit index of at least 0.85) and more stringent correlation thresholds can lead to networks being closer to a scale-free network (i.e. to higher values of R2). We give now these dependencies for the three networks in new Fig. 1D. Please note that reducing by a threshold on significance of correlation is a different reduction method and it is neither expected nor intended that the resulting networks are scale-free. In order to illustrate the relationship between the two reduction methods, we introduced new illustrations in supplemental figure S1. Therein, - we show in new panel S1A that for the metabolite and mRNA layer, due to the datasets not having NA values and thus the sample size being the same for each and every pair of metabolite or mRNA species, the p-values of the correlation being different from zero depend monotonously on the absolute correlation values. This is not the case for the protein dataset for which different sample sizes may occur. Therefore, the correlation value for some interactions will be backed by fewer data points only which leads to increased p-value despite the same correlation value. - we show in new S1B that, for the mRNA and metabolite data layer, the scale-free fit index for networks reduced by different significance thresholds can be directly compared to those from Fig. 1D. - we show in new S1C how the significance threshold of correlation being different from zero relates to the scale-free score for the protein dataset and where the four networks reduced by the four considered significance thresholds locate therein. >To demonstrate that SBM can provide biological insight, more detailed analysis on the benchmark data sets could be useful. For example, what are unique and common pathways for different breast cancer subtypes, how well are these findings consistent with existing knowledge? Response: We agree that a comparison of results between breast cancer subgroups to provide biological insights into the utility of the clustering would be useful. However, our work focusses on determining whether SBMs are suitable to represent biomolecular networks as they have not been explored for molecular network analysis. We indeed expect additional and complementary biological findings to what is already known with the SBM approach. To strengthen our work, we also added some showcase examples of biological interpretation (new Figs. 4, 5, Table 2 and Description in the text). They show that the SBM-derived clustering can detect biological pathways known to be implicated in breast cancer according to oncogenic signatures from MSigDB, such as extracellular matrix organization or the cell cycle (see new Fig. 4), or fatty acid biosynthesis (Table 2). In addition, the relevance of further processes can be predicted, e.g. the chromatin organization (new Fig. 5). >Figure 4 shows SBM based confidence score can be used to predict the existence of an edge. However, it is not clearly stated how the edges are determined as putatively missing or spurious. Also, the description for figure 4A is hard to understand. Response: We extended the description of the edge detection by adding the following sentences in the section ‘Assessing edge relevance by SBM-based edge confidence scores’: ‘For the six reduced networks, all edges that exist in the network were considered as ‘putatively spurious’. Similarly, all edges that were not in the network because they had been removed from the (fully connected) correlation-based network during the reduction procedure were considered as ‘putatively missing’.’ We clarified the legend to new Fig. 6A (old Fig. 4A) that now reads: ‘(A) Histograms of edge confidence scores for the best fitting SBM of the three networks reduced by criterion on scale-freeness. Top: spurious edge confidence scores, computed for all edges existing in the network; middle and bottom: missing edge confidence scores, computed for all edges that were removed during the reduction procedure (middle: missing edges between nodes with degree>0 in the reduced network, bottom: missing edges adjacent to a node of degree zero in the reduced network).’" } ] }, { "id": "47223", "date": "29 Apr 2019", "name": "Silvia Pineda San Juan", "expertise": [ "Reviewer Expertise Statistics", "Computational biology", "Data analysis", "Genomics." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript provides an innovative way of using stochastic block models to perform networks for different omics data. The idea is interesting but the application should be better characterized. I was expecting a network integrating the different datasets, but the network is performed separately per each dataset. This fact should be addressed at the very beginning to avoid confusion to the readers. Also, the use of the three datasets is not justified, why these three and no others? It looks like the authors want to compare the three networks applied to the three datasets (mRNA, proteins, and metabolites) but the different sample size among other factors makes them not comparable. Finally, one of my main concern is that there is not a clear conclusion of the study, are they proposing a network that is better than the ones that already exist? (this is not assessed). Are they finding new biological insights for breast cancer? (this is not shown). The final conclusion is not clear since the authors did not give a biological example of the method application. It is not clear if the method outperforms others or if the method is able to find new biological interpretations, etc. Why people should use this method? What type of information will they obtain?\n\nMore specific points:\n\nData preparation and network generation:\nIt is not clear to me why the authors used mRNA, protein and metabolites for this network study. Is it just because they were available, or is there a hypothesis under this selection? Why don't they use other omics data available in the TCGA? The sample size is pretty reduced for the proteomics data with only 36 samples. The other thing I do not understand is why they used metabolomics data measured in other individuals. And for the mRNA they do not give an exact sample size.\n\nWhy do they use Bonferroni for mRNA expression and Benjamini-Hochberg for protein, metabolite and just 0.01 for protein? This should be better justified.\n\nThey replace 0 values by NA, why? There is a big difference between a lack of expression and a missing value.\n\nDo they filter for those genes that have a very low expression among samples? They only specify this for protein data, but what about mRNA expression? Are they considered only 0 for low expression or a very small cut-off normally used in mRNA analysis?\n\nRegarding the scale-free reduction. They explain that the technique removes weak links until met a criterion based on WGCNA package, but it does not well explain the process for this and how they applied this to the data. Please explain.\n\nFitting SBM:\nThey built the network based on a stochastic block model (SBM) representation, the nodes of a network are partitioned into blocks according to their similarity in connectivity. It is not clear how the SBM is applied to the data and how the similarities are obtained.\n\nFor the SBM representing biological function:\nThe whole module is unclear to me since they do not provide any biological or functional interpretation. I don't understand the goal of this. Figure 3 is also quite confusing.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4832", "date": "27 Aug 2019", "name": "Katharina Baum", "role": "Author Response", "response": "We thank Dr. Silvia Pineda San Juan for the careful revision of our manuscript. Her comments helped us enhancing the clarity and stringency of our presentation and the reader's accessibility to our work. We respond to her specific comments and describe the introduced changes to version 2 of our manuscript below. >The manuscript provides an innovative way of using stochastic block models to perform networks for different omics data. The idea is interesting but the application should be better characterized. I was expecting a network integrating the different datasets, but the network is performed separately per each dataset. This fact should be addressed at the very beginning to avoid confusion to the readers. Response: We thank the reviewer for her appreciation of our idea. In order to improve the clarity of the description of the data usage from the beginning, we reformulated the corresponding sentences in the abstract and introduction which now read: ‘We apply SBM-based analysis independently to three correlation-based networks of breast cancer data originating from high-throughput measurements of different molecular layers: either transcriptomics, proteomics, or metabolomics.’ ‘Here we showcase the SBM-based analysis (overview in Fig. 1A) for three networks of different molecular types, derived from either transcriptomic, proteomic or metabolomics data of breast cancer tumours.’ >Also, the use of the three datasets is not justified, why these three and no others? It looks like the authors want to compare the three networks applied to the three datasets (mRNA, proteins, and metabolites) but the different sample size among other factors makes them not comparable. Finally, one of my main concern is that there is not a clear conclusion of the study, are they proposing a network that is better than the ones that already exist? (this is not assessed). Are they finding new biological insights for breast cancer? (this is not shown). The final conclusion is not clear since the authors did not give a biological example of the method application. It is not clear if the method outperforms others or if the method is able to find new biological interpretations, etc. Why people should use this method? What type of information will they obtain? Response: In order to sharpen the focus, we clarified in the abstract and at the end of the introduction that our goal is to pave the ground for the usage of the SBM model for different types of biomolecular networks. Investigating the capability of SBMs for representing and analysing different types of biological networks was the key challenge addressed in our article. We did not intend to compare the networks between layers, but rather assess to which extent the SBM is applicable to derive useful information in terms of (i) relevant clustering as well as (ii) network-based, alternative edge scores. We have shown that a lot more SBM-predicted blocks have biological counterparts (i.e. more genes or metabolites associated with certain Reactome or KEGG terms are clustered together) than expected by chance, and biological processes known to be relevant to the examined phenotype are can be derived (new Fig. 4, 5, Table 2). In addition, we showed that the SBM-based edge relevance scores coincide with the correlation values (which have not been given to fit the network to the SBM). These results support our hypothesis that the SBM is suitable to represent and analyze biomolecular networks in which interactions are derived from correlations. This opens the avenue to new types of analyses using the SBM for which this work lays the foundation. To strengthen our findings, as suggested by the reviewer, we now also include more biological interpretations of the clustering results (see Figs. 4, 5, Table 2). They show that the SBM-derived clustering can detect biological pathways known to be implicated in breast cancer according to oncogenic signatures from MSigDB, such as extracellular matrix organization and the cell cycle (see new Fig. 4), or fatty acid biosynthesis (Table 2). In addition, the relevance of further processes can be predicted, e.g. the chromatin organization (new Fig. 5). >More specific points: >Data preparation and network generation: It is not clear to me why the authors used mRNA, protein and metabolites for this network study. Is it just because they were available, or is there a hypothesis under this selection? Why don't they use other omics data available in the TCGA? The sample size is pretty reduced for the proteomics data with only 36 samples. The other thing I do not understand is why they used metabolomics data measured in other individuals. And for the mRNA they do not give an exact sample size. Response: Protein, mRNA and metabolites are key molecules in cells, which are widely applied, in isolation or in combinations, in different biomedical research domains. Their abundance and interconnectedness in networks are therefore of high interest if aiming to characterize cells or tumorous tissue. We used them to illustrate three examples for cellular interaction networks, whose interactions have different biological meanings. Moreover, in the revised article now we explain that we use these networks to reflect different characteristics of tumorous tissue/cells. Other potential data types could be e.g. mutations, copy number variation, DNA methylation or miRNAs, which are interesting avenues to further explore. While they could be useful, there are some caveats associated with them, e.g.: the derived interactions within layers are even less directly interpretable than for mRNA, protein or metabolite; networks generated from mutations and copy number variation are extremely sparse; the functional interpretation of DNA methylation data relies on transcription and the resulting networks are extremely big; the roles of miRNAs are less well known. Therefore, we decided to restrict our analyses to the three biomolecular entities mRNA, proteins and metabolites. We included these considerations when introducing the employed data layers in the results part of the revised manuscript. Concerning the metabolite data: We are convinced that metabolites characterize a highly interesting layer of intra-tumor processes which is complementary to the gene expression associated layers mRNA and proteins. Unfortunately, metabolomics have not been measured for TCGA samples which is why we resorted to an alternative cohort for this data layer. Please note that the number of tumour samples for each data layer were stated in the methods section (mRNA: 237, protein: 36, metabolite: 68). We now included that value for the mRNA data layer also into the main text. >Why do they use Bonferroni for mRNA expression and Benjamini-Hochberg for protein, metabolite and just 0.01 for protein? This should be better justified. Response: We explain our approach for network reduction more in detail in the methods and results part of the revised version of the manuscript. In fact, at first, we applied both Bonferroni and Benjamini-Hochberg correction methods, both of which are widely used and accepted, along with the classical significance thresholds 0.01 and 0.05 to the networks of all three data layers (see Fig. 1C). We finally chose the correction method and threshold for each data layer considering a trade-off between minimal network size (i.e. minimal computation time for the subsequent fit to SBM) and maximal connectedness of the reduced network: We used the combination which provided a high degree of reduction (less than 10 million edges in the network) while maximizing the size of the largest connected component in the network. While the stringent Bonferroni correction is necessary to achieve a sufficient degree of reduction for the mRNA network, it severely disrupted the connectedness for the protein and metabolite data layer leading to less than 30% or 65% of the nodes being in the largest connected component for protein or metabolite, respectively (see Fig 1C). Using different correction methods and significance thresholds serve as examples of typical scenarios which could be envisioned during network reduction. Finally, every user could use their own thresholds reasonable for network reduction. >They replace 0 values by NA, why? There is a big difference between a lack of expression and a missing value. Response: We replaced NAs by -10 in the log-counts of the RNAseq data. As stated in the methods part, the reason for this is the following: The RNAseq data are logarithmized (relative) counts. In this case the NAs are therefore artefacts from logarithmizing a zero count. Our replacement served to reverse this artefact. >Do they filter for those genes that have a very low expression among samples? They only specify this for protein data, but what about mRNA expression? Are they considered only 0 for low expression or a very small cut-off normally used in mRNA analysis? Response: Apart from the replacement of NAs by very small values (-10 in log-counts) in the mRNA data, we use the data as provided by TCGA. The rationale for this is to reduce bias with respect to which nodes we consider. Since we do not focus on expression strength, but on the connection between molecular species, i.e. co-variation of expression, also lowly abundant species could, and indeed do, play a role, i.e. they have non-zero degree in the reduced networks. We dedicate new Figure S1D to an illustration of this fact. One can imagine multiple other criteria of entity removal prior to analysis (e.g. tissue-specific GTEx, using pathways of interest) but this is not the focus of our work. >Regarding the scale-free reduction. They explain that the technique removes weak links until met a criterion based on WGCNA package, but it does not well explain the process for this and how they applied this to the data. Please explain. Response: We incorporated a more detailed explanation of this reduction technique in the methods part and added an additional panel to Fig. 1 (Fig. 1D) which illustrates intermediate results of the process. This is the revised methods part: ‘For the reduction by imposing a scale-free architecture of the reduced network, we employed the pickHardThreshold function of the WGCNA package (Langfelder et al, 2008) with the default requirement (0.85) on goodness of fit to a power-law degree distribution of the nodes. Given the symmetric absolute correlation matrix of the network edges, this function reduces the network by one of a given set of edge thresholds at a time and determines the scale-free fit index R2 which lies between 0 (bad fit) and 1 (perfect fit) by comparing the resulting degree distribution of the reduced network to a power-law degree distribution. The lowest of the tested edge thresholds that gives a scale-free fit index > 0.85 is reported as estimated threshold. For the edge thresholds, we started with a grid with stepsize 0.05 between 0.3 and 0.95, refining according to the resulting estimates to vectors with stepsize 0.001 between 0.5 and 0.625 for the mRNA network, between 0.7 and 0.82 for the protein network and between 0.3 and 0.4 for the metabolite network. Finally estimated edge correlation thresholds were 0.603 (mRNA), 0.788 (protein), and 0.375 (metabolite).’ >Fitting SBM: They built the network based on a stochastic block model (SBM) representation, the nodes of a network are partitioned into blocks according to their similarity in connectivity. It is not clear how the SBM is applied to the data and how the similarities are obtained. Response: We extended the corresponding section describing the SBM in the Methods, in which we now also summarize the equations underlying the building of the SBM and the relationship between model likelihood and properties of the network graph derived from the data. >For the SBM representing biological function: The whole module is unclear to me since they do not provide any biological or functional interpretation. I don't understand the goal of this. Figure 3 is also quite confusing. Response: The assumption behind the analysis is that a predicted block structure is meaningful if many of the predicted blocks have a biological counterpart, e.g., biological pathway. Such an analysis was performed as a starting point for assessing potential biological relevance. Nevertheless, because we agree that additional biological interpretations would benefit the reader, we included other examples of biological interpretations (new Figs. 4, 5, Table 2 and accompanying description in the results section). In addition, in order to improve accessibility, we reduced the contents of Fig. 3 (by only showing results for one distance measure) and clearly indicated which SBM hierarchy level the examined blocks stem from." } ] }, { "id": "47834", "date": "05 Jun 2019", "name": "Ana Conesa", "expertise": [ "Reviewer Expertise Bioinformatics", "functional genomics", "transcriptomics" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript describes the application of stochastic block models (SBMs) as a method to construct biological networks from different omics data. Starting from correlation-based networks of breast cancer data, SBM delivers different modules (or blocks) of the network. In order to assess the usefulness of this approach, the authors evaluate the biological meaning of the modules obtained performing a functional enrichment analysis for each module of the network. Finally, the authors include the edge confidence score computation for each edge in the resulting network.\n\nThis novel application of SBMs with omics data is interesting but its utility is not clearly explained in the manuscript. First, the title is confusing, since I understood a single network is obtained from different omics data, i.e. a multiomic integration analysis, however it is not performed. Furthermore, why are these three datasets used? It seems the authors want to evaluate the SBM-based networks starting from datasets with different complexity (different number of features/samples), but the reason is not properly explained. As for the SBM model definition, because of the microcanonical formulation because of its hard constraints, network underfitting could be an important issue. Although underfitting is solved using nested (hierarchical) or degree-corrected SBMs, an explanation of this phenomena and how to deal with it would be suitable in “Fit to SBM” methodological section. Regarding the functional enrichment, the choice of the pathway database is not clear. Why is Reactome chosen for genes annotation and KEGG for metabolites? Taking into account that breast cancer related biological findings is not the scope of this manuscript, a deeper analysis of biological information revealed by network blocks is necessary: which pathways are block-specific? Which pathway or pathways are present in the bigger block? Do the pathway distances between blocks have any relation with previous knowledge about breast cancer? Furthermore, the pathway distances within and between blocks are computed just for the lower level of blocks. Since the network is hierarchical, pathway distances can be obtained for each level. This would indicate if the hierarchy is biological meaningful too. Moreover, the same genes are known to be annotated to related pathways. Therefore, this expected result does not represent a validation to determine the biological meaning of the blocks in the network. A good approach could be to select a gene X and see if the rest of the genes of the pathways, where gene X is annotated, are in the same block of gene X (or in near blocks in the hierarchy). And repeat this for a high number of genes. Finally, the edge confidence score looks like important to improve the network characterization, however they do not use this score to optimize the final network. I was expecting the comparison between the original SBM-based network and the corrected SBM-based network via edge confidence score optimization. In order to use the edge confidence scores, a way to compute thresholds for missing and spurious edges should be proposed.\n\nResuming, SBM-based biological networks are a new way to represent omics data and it represents a novel approach for SBMs. However, a novel application should be coupled with a suitable interpretation of its results. Moreover, the main conclusion or this manuscript is not clear and the main question for a new methodological development is unanswered, why should I use this method?\n\nSpecific points:\n1) Data preprocessing:\nWhy are the three datasets used? Is it just a matter of different complexity? Datasets are not completely characterized as the number of features is not indicated. Authors say missing values of metabolite data were imputed previously. It is not clear whether they performed the imputation or not. In case they did, the algorithm used is not indicated. If missing values of metabolite data were imputed, why weren't the missing values of protein data?\n\n2) Network generation and reduction:\nThey use different test-multiple correction rates for mRNA, protein and metabolites in order to obtain a similar degree of reduction. The authors should demonstrate this does not lead to any biases. The reduction objective is set at 107 edges, is there any reason? In Figure 1.D, a linear relationship is expected for the blue dots but it not perfectly linear, why? Could the scale-free reduction be modified to improve it?\n\n3) Stochastic block models:\nHow are the number of initial partitions (500) defined? Is there any relation between the suitable number of initial partitions and the network complexity (size)?\n\n4) Functional enrichment:\nThey use Reactome database for genes annotation and KEGG for metabolites, but there is no an explanation of why that choice. Figure 3.C is confusing. I do not understand the objective of this analysis. In order to demonstrate the biological meaning of the hierarchical SBM, the pathway distances comparison within and between blocks at every hierarchical level (as in Figure 3.B) would be better than the provided analysis. As mentioned before, a deeper biological interpretation of the network is necessary. I am not sure what the clustering analysis is contributing to. Modules obtained by WGCNA shows good results looking at the number significant of pathways, but are they biologically relevant? Are they related to breast cancer?\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4833", "date": "27 Aug 2019", "name": "Katharina Baum", "role": "Author Response", "response": "We thank Dr. Ana Conesa and Manuel Ugidos for their time and effort they invested into commenting our work. They raised important points that we carefully addressed in our revised version 2 of the manuscript. Please find our specific replies and the description of introduced changes below. >This manuscript describes the application of stochastic block models (SBMs) as a method to construct biological networks from different omics data. Starting from correlation-based networks of breast cancer data, SBM delivers different modules (or blocks) of the network. In order to assess the usefulness of this approach, the authors evaluate the biological meaning of the modules obtained performing a functional enrichment analysis for each module of the network. Finally, the authors include the edge confidence score computation for each edge in the resulting network. >This novel application of SBMs with omics data is interesting but its utility is not clearly explained in the manuscript. First, the title is confusing, since I understood a single network is obtained from different omics data, i.e. a multiomic integration analysis, however it is not performed. >Furthermore, why are these three datasets used? It seems the authors want to evaluate the SBM-based networks starting from datasets with different complexity (different number of features/samples), but the reason is not properly explained. >As for the SBM model definition, because of the microcanonical formulation because of its hard constraints, network underfitting could be an important issue. Although underfitting is solved using nested (hierarchical) or degree-corrected SBMs, an explanation of this phenomena and how to deal with it would be suitable in ‘Fit to SBM’ methodological section. >Regarding the functional enrichment, the choice of the pathway database is not clear. Why is Reactome chosen for genes annotation and KEGG for metabolites? >Taking into account that breast cancer related biological findings is not the scope of this manuscript, a deeper analysis of biological information revealed by network blocks is necessary: which pathways are block-specific? Which pathway or pathways are present in the bigger block? Do the pathway distances between blocks have any relation with previous knowledge about breast cancer? > Furthermore, the pathway distances within and between blocks are computed just for the lower level of blocks. Since the network is hierarchical, pathway distances can be obtained for each level. This would indicate if the hierarchy is biological meaningful too. Moreover, the same genes are known to be annotated to related pathways. Therefore, this expected result does not represent a validation to determine the biological meaning of the blocks in the network. A good approach could be to select a gene X and see if the rest of the genes of the pathways, where gene X is annotated, are in the same block of gene X (or in near blocks in the hierarchy). And repeat this for a high number of genes. >Finally, the edge confidence score looks like important to improve the network characterization, however they do not use this score to optimize the final network. I was expecting the comparison between the original SBM-based network and the corrected SBM-based network via edge confidence score optimization. In order to use the edge confidence scores, a way to compute thresholds for missing and spurious edges should be proposed. > Resuming, SBM-based biological networks are a new way to represent omics data and it represents a novel approach for SBMs. However, a novel application should be coupled with a suitable interpretation of its results. Moreover, the main conclusion or this manuscript is not clear and the main question for a new methodological development is unanswered, why should I use this method? Response: We thank the reviewers for the nice summary of our work. In order to sharpen the focus, we clarified in the abstract and at the end of the introduction that our goal is to pave the ground for the usage of the SBM model for different types of biomolecular networks. Investigating the capability of SBMs for representing and analysing different types of biological networks was the key challenge addressed in our article. We did not intend to compare the networks between layers, but rather assess to which extent the SBM is applicable to derive useful information in terms of (i) relevant clustering as well as (ii) network-based, alternative edge scores. We have shown that a lot more SBM-predicted blocks have biological counterparts (i.e. more genes or metabolites associated with certain Reactome or KEGG terms are clustered together) than expected by chance, and we show in our revised version that biological processes known to be relevant to the examined phenotype can be derived (new Figs. 4, 5, Table 2). In addition, we showed that the SBM-based edge relevance scores coincide with the correlation values (which have not been given to fit the network to the SBM). These results support our hypothesis that the SBM is suitable to represent and analyze biomolecular networks in which interactions are derived from correlations. This opens the avenue to new types of analyses using the SBM and its output for which this work lays the foundation. We hope that the reviewers will also find that the revised version supports this reasoning and the derived conclusion. In response to other points that were raised only here but not in the specific points include below: To improve the clarity of the description of the data usage from the beginning, we reformulated the according sentences in the abstract and introduction which now read: ‘We apply SBM-based analysis independently to three correlation-based networks of breast cancer data originating from high-throughput measurements of different molecular layers: transcriptomics, proteomics, or metabolomics.’ ‘Here we showcase the SBM-based analysis (overview in Fig. 1A) for three networks of different molecular types, derived from either transcriptomic, proteomic or metabolomics data of breast cancer tumours.’ In order to facilitate the understanding of the different SBM versions, we added an explanation on the hierarchical and degree-corrected version of the SBM into the methods part and comment on the relationship between underfitting and the hierarchical SBM in the revised methods section. >Specific points: >1) Data preprocessing: Why are the three datasets used? Is it just a matter of different complexity? Datasets are not completely characterized as the number of features is not indicated. Response: Protein, mRNA and metabolites are key molecules in cells, which are widely applied, in isolation or in combinations, in different biomedical research domains. Their abundance and interconnectedness in networks are therefore of high interest if aiming to characterize cells or tumorous tissue. We used them to illustrate three examples for cellular interaction networks, whose interactions have different biological meanings. Moreover, in the revised article now we explain that we use these networks to reflect different characteristics of tumorous tissue/cells. The number of features (i.e. number of mRNAs, proteins and metabolites) are given in the network characteristics in Table 1 (‘entities’ before NA removal and afterwards). In order to improve accessibility to the reader, we also state these numbers in the methods part of the revised manuscript and clearly identify them as feature count (‘mRNA, protein, metabolite data for ER-breast cancer tumors’). >Authors say missing values of metabolite data were imputed previously. It is not clear whether they performed the imputation or not. In case they did, the algorithm used is not indicated. If missing values of metabolite data were imputed, why weren't the missing values of protein data? Response: We used the metabolite data as provided in Budczies et al. 2009 in which imputation had been performed. In order to clarify this, we reformulated the sentence in the methods part to ‘The metabolite data did not contain missing values as imputation had been performed in the original publication[Budczies et al. 2009].’ We used the processed protein data as provided in the original publication (Mertins et al, 2014). In general, in contrast to missing values in metabolite data which are usually considered to occur due to abundance below detection limit, missing values in protein data can have multiple different reasons making imputation less straightforward. Because we do not focus on data pre-processing here, we kept the data as close as possible to those originally published and rather took the opportunity to provide an idea of how to incorporate datasets with missing data into our proposed analysis framework. >2) Network generation and reduction: They use different test-multiple correction rates for mRNA, protein and metabolites in order to obtain a similar degree of reduction. The authors should demonstrate this does not lead to any biases. The reduction objective is set at 107edges, is there any reason? Response: We reformulated and describe this part in more detail in the revised manuscript. In fact, we intended to provide examples for different corrections methods and thresholds which are frequently used, and to achieve a high degree of reduction (to reduce runtime for SBM fit) while still ensuring that the resulting biomolecular network is as connected as possible (see Fig. 1C). Similarity of the degree of reduction between the three data layers was not a goal to achieve. The border of 107 edges arose out of computation time considerations (runtime scales with edge count for SBM fit in graph-tool, they correspond to several hours and Gbs of memory consumption for a single initialization), and thus could be adapted to each user’s case. >In Figure 1.D, a linear relationship is expected for the blue dots but it not perfectly linear, why? Could the scale-free reduction be modified to improve it? Response: Usually, a perfect fit to scale-freeness cannot be achieved because altering the cut-off threshold leads to removal of multiple edges at the same time; the achievable combinations of which edges belong to the network are fully determined by the correlation values associated to the edges. We show possible scale-free fit indices (that index is a measure of how closely the node degree distribution resembles a power-law, i.e. how close the network is to scale-freeness) of the networks reduced by edge thresholding in new Fig. 1D. Therein, it also becomes clear that other thresholds could be used and can lead to ‘better’ scale-freeness in the resulting networks – while at the same time making the resulting network less connected. Other specific requirements on the degree of scale-freeness can be adapted on a case-by-case basis. >3) Stochastic block models: How are the number of initial partitions (500) defined? Is there any relation between the suitable number of initial partitions and the network complexity (size)? Response: The number of initializations was chosen as first assessment whether this is sufficient to be able to distinguish between the different SBM types – which was the case. As is clear, for some networks different SBM types are more closely related whereas for others, they are well more separated. This will depend on the actual network/data that is fitted and serves as orientation. Of course, more initializations are always better, but it is subject to a trade-off between computation time and finding a good partition. >4) Functional enrichment: They use Reactome database for genes annotation and KEGG for metabolites, but there is no an explanation of why that choice. Response: We added the explanation to the revised version. Reactome provides a hierarchical annotation which qualifies it for comparison to the hierarchical structure given by the SBM, KEGG is one of the annotation databases used most for metabolites. >Figure 3.C is confusing. I do not understand the objective of this analysis. In order to demonstrate the biological meaning of the hierarchical SBM, the pathway distances comparison within and between blocks at every hierarchical level (as in Figure 3.B) would be better than the provided analysis. Response: We altered the accompanying description of Fig. 3C. What we intended is to relate the distance of two SBM blocks in the SBM hierarchy (x-axis) to the distances between the Reactome terms associated to the blocks (y-axis). For a matching hierarchy structure of Reactome and SBM, we would expect a positive correlation, i.e. more distant SBM blocks in the SBM hierarchy having more distantly related Reactome terms. We do not find convincing evidence for this. To clarify the approach, we restructured Fig. 3 (by moving the results for the second distance measure into the supplement, new Fig. S3). In addition, we now also include the analysis of within-block distance vs. between-block distances for the other hierarchy levels as suggested by the reviewer (new Fig. 3D). As in our previous analysis in Fig. 3C, we cannot find strong evidence for the hierarchy of the SBM coinciding with the Reactome hierarchy. Please note that the numbers of blocks with annotation get low in higher hierarchy levels, thereby reducing the number of hierarchy levels that can be considered for this analysis. >As mentioned before, a deeper biological interpretation of the network is necessary. I am not sure what the clustering analysis is contributing to. Modules obtained by WGCNA shows good results looking at the number significant of pathways, but are they biologically relevant? Are they related to breast cancer? Response: We provide additional biological interpretations of the clustering results in the revised version of the manuscript (see new Figs. 4, 5, Table 2). They show that the SBM-derived clustering can detect biological pathways known to be implicated in breast cancer according to oncogenic signatures from MSigDB, such as extracellular matrix organization or the cell cycle (see new Fig. 4), or fatty acid biosynthesis (Table 2). In addition, the relevance of further processes can be predicted, e.g. chromatin organization (new Fig. 5)." } ] } ]	1	https://f1000research.com/articles/8-465
https://f1000research.com/articles/8-1491/v1	23 Aug 19	{ "type": "Research Article", "title": "Evaluation of fracture resistance of Cerasmart and lithium disilicate ceramic veneers with different incisal preparation designs: an in vitro study", "authors": [ "Bushra Mohammed", "Jylan EL-Guindy", "Jylan EL-Guindy" ], "abstract": "Background: Cerasmart hybrid material offers specific advantages such as less fragility and more flexibility than glass ceramics. This material also has the option of readily modifying or repairing the surface and favorable stress-absorbing characteristics. In our study, Cerasmart hybrid and lithium disilicate ceramic laminate veneers with two different preparation designs were compared with regards to their fracture resistance. Methods: A total of 52 of comparable human central maxillary incisors were used. Group A (n=26) was made up of Cerasmart hybrid ceramic laminate veneers were fabricated from Cerasmart blocks, while Group B (n=26) was made up of lithium disilicate ceramic laminate veneers were made of IPS e.max pressable ingots. Each group was subdivided in two equal subgroups according to preparation designs. Subgroup I comprised Featheredge preparation design and subgroup II: Wraparound preparation design. All samples were subjected to thermocycling between 5°C and 55°C in a water bath for a total of 1750 cycle with 10 seconds dwell time at each bath. The fracture load strength test was performed using a universal testing machine. Results: There was no statistically significant difference between all groups. E.max wraparound group recorded the highest fracture resistance mean value (422.1 N) followed by Cerasmart wraparound group (317.23 N), then e.max featheredge group (289.6 N), and finally Cerasmart featheredge group (259.3 N) had the lowest value as analyzed by one-way ANOVA. Conclusions: The Cerasmart hybrid material could be considered as a valid alternative to the widely used IPS e.max material. The fracture resistance of laminate veneers is not influenced by different type of preparation designs.", "keywords": [ "Laminate veneers", "Lithium disilicate ceramic", "Hybrid ceramic", "Fracture resistance", "Preparation designs." ], "content": "Introduction\n\nLaminate veneer restorations have gained popularity and patient contentment owing to their good esthetics and highly conservative tooth preparation designs1. Pressable and machinable ceramics have developed as an option to the conventional porcelain layering manufacturing method2. Ceramics have the benefits of elevated flexural strength and stability of color, while their drawbacks include elevated antagonistic tooth wear and less conservative tooth preparation3. Composite resins can overcome these two drawbacks, although wear of the material itself is higher4. While ceramics have higher stiffness and hardness values than the natural tooth, reduced values are shown by composite resins. Therefore, a material is needed that combines the benefits of ceramics with those of composites, causes minimal wear of both the material itself and the antagonistic tooth and preserves the structure of the sound tooth3.\n\nThe success rate of ceramic laminate veneers has been clinically evaluated and has shown range from 18 months up to 15 years; the rate of success varies between 75% and 100%. De-bonding, fracture and micro-leakage failures are seen in ceramic veneers4. Fractures accounted for 67% of the total failures over a 15-year clinical performance period5.\n\nThis research aimed to compare the impact of hybrid and lithium disilicate ceramic laminate veneer on their resistance to fracture with two distinct preparation configurations after thermal cycling.\n\n\nMethods\n\nThis study was approved by the Research Ethics Committee of the Faculty of Dentistry, Cairo University. Approval number 15531.\n\nA total of 52 human central maxillary incisors were selected for the present study. Teeth scaling and polishing was done to remove any remnants, then they were stored in saline solution at room temperature. To facilitate the handling of teeth, the root of each tooth was mounted vertically with its long axis in Epoxy resin blocks (CMB, Egypt).\n\nTeeth were divided into two equal groups (n=26) according to material: Group A (Cera), Cerasmart laminate veneers, and Group B (e.max), IPS e.max Press laminate veneers). These groups were further subdivided into two subgroups according to preparation design (subgroup I, Featheredge preparation design, and subgroup II, Wraparound preparation design).\n\nStandardized teeth preparation was done using a five-axis computer numerically controlled (CNC) milling machine (Centroid M400 CNC, USA) with water coolants. The labial surface preparation was performed in mesio-distal direction in two planes (cervical one third and incisal two thirds) (Figure 1). The incisal reduction was only performed in wrap-around design (Figure 2). The parameters are summarized in Table 1.\n\nCerasmart laminate veneers were fabricated using a CAD/CAM system (Omnicam, CEREC MC XL SW 4.0; Sirona Dental Systems GmbH, Germany) using Cerasmart blocks (GC Corporation, Tokyo, Japan). The spacer was adjusted at 30 μm and laminate veneer thickness at 0.5 mm cervically and 0.8 mm incisally (Figure 3, Figure 4).\n\n(a) Cervical, (b) middle, (c) incisal.\n\nIPS e.max Press laminate veneers were fabricated in two steps; first, digital waxing up automated by Exocad CAD software (Exocad GmbH, Germany) to overcome the variation of manually fabricated restorations and to standardize the thickness of all samples. Second, the pressing procedure was done using IPS e.max Press ingots (Ivoclar Vivadent AG, Principality of Liechtenstein) according to the manufacturer’s instructions.\n\nFinally, all laminate veneers were checked on the corresponding prepared tooth for proper seating and marginal adaptation (Figure 5, Figure 6).\n\nAll restorations were cleaned using an ultrasonic cleaner (Shenzhen, China) for 180 seconds, the fitting surfaces of the veneers were etched with 9.5% hydrofluoric acid gel IPS Ceramic Refill (Ivoclar Vivadent AG, Principality of Liechtenstein) for 60 seconds for Group A teeth and for 20 seconds for Group B teeth, followed by thorough washing by air/water spray for 30 seconds and drying using air spray. Silane coupling agent Monobond-S (Ivoclar Vivadent AG, Principality of Liechtenstein) was applied for 60 seconds and air-dried before cementation.\n\nThe prepared teeth were acid etched using 37% phosphoric acid Scotchbond Universal Etchant (3M ESPE, USA) for 15 seconds and rinsed with air/water spray for another 10 seconds then dried with air spraying. With a fully saturated micro-brush tip two consecutive coats of Single Bond Universal adhesive (3M ESPE, USA) were applied on the prepared enamel surface and thinned gently by air spray for 2–5 seconds. Light cure RelyX Veneer cement (3M ESPE, USA) was used to lute the laminate veneers on their corresponding prepared teeth. The cement was applied on inner surface of the veneers and the veneers were seated with gentle finger pressure on the prepared teeth. Excess cement was removed immediately with an explorer, the exposed margins were covered with glycerin gel to prevent formation of an oxygen-inhibiting layer and ensure the complete polymerization of the cement. Margins were light cured for 20 seconds from the incisal, lingual, mesial and distal directions each respectively.\n\nAll samples were subjected to thermocycling between 5° and 55°C in a water bath at each temperature for a total of 1750 cycles with a dwell time of 30 seconds at each temperature (water bath) and 10 seconds transport time between the two baths. Fracture resistance test of all samples was performed using a universal testing machine (Instron, Model 3345; Instron industrial, USA) by compressive mode of load applied at 135° angle to the lingual surface of the tooth to simulate the clinical situation as closely as possible6. The load was applied using a metallic rod with flat tip (3.8 mm diameter) attached to the upper movable compartment of testing machine traveling at cross-head speed of 1 mm/min with tin foil sheet in-between to achieve homogenous stress distribution and minimization of the transmission of local force peaks (Figure 7). The load required to fracture was recorded in Newtons.\n\nData were gathered, tabulated and analyzed statistically using SPSS statistical software (Version 21, Chicago, IL, USA). One-way ANOVA followed by pair-wise Tukey’s post-hoc tests were performed to detect significance between groups. Student t-test was performed to detect significance between paired groups. The level of significance was set at 5% for all statistical analyses and confidence interval at 95%.\n\n\nResults\n\nE.max wrap-around group recorded statistically non-significant (P>0.05) highest fracture resistance mean value (422.1 N) followed by the Cera wrap-around group (317.23 N), and then E.max feather-edge group (289.6 N), while Cera feather-edge group recorded the lowest fracture resistance mean value (259.3 N) (Table 2 and Figure 8).\n\nSD, standard deviation; NS, not significant.\n\nThe e.max groups recorded statistically non-significant (P>0.05) higher fracture resistance mean values than Cera groups. Wrap-around groups recorded statistically non-significant higher fracture resistance mean values than feather-edge groups as indicated by the results of Student’s t-test.\n\nThe following aspects were considered to assess the failure:\n\n1- Tooth (Intact or fractured)\n\n2- Veneer (Intact or fractured)\n\n3- Tooth-veneer junction (Intact or debonding)\n\nThe behavior of samples differed; the results are tabulated in Table 3. None of the restorations tested were fractured. The fracture when occurred was in the tooth structure; root, cervical or incisal edge fracture. The type of failure was either cohesive; fracture in the tooth, adhesive; debonding of the veneer without fracture of the tooth, or adhesive-cohesive; debonding of the veneer with fracture of the tooth (Figure 9, Figure 10). E.max groups showed cohesive type of failure only. Cera groups showed all types of failure.\n\n(a) Intact tooth (adhesive failure); (b) fractured incisal edge (adhesive-cohesive failure).\n\n\nDiscussion\n\nCeramic laminate veneers are a popular, safe and successful technique to restore discolored, worn, malformed or broken teeth7. The ongoing development of esthetic and functional ceramic adhesive restorations allows the patient’s smile and self-esteem to be improved8.\n\nFor this research, maxillary incisors were chosen as more fractures are observed in veneers prepared on maxillary incisors and due to the elevated demand for aesthetics in this area9.\n\nAlthough some investigators10,11 used periodontal ligament simulation material, it was not used in this research so that the gradual load applied to the coronal part of the embedded tooth would not have been lightened by the interposition of the simulation material between the tooth root and the surrounding epoxy resin12.\n\nDifferent designs of incisal edge teeth preparations for laminate veneers were suggested by multiple authors6–9. Meijering et al.13 found that the incisal edge preparation design was not linked to the restoration success in a 2.5-year clinical study. Prasanth and Friedman14 found that both feather-edge and butt joint preparations were superior to the palatal chamfer. Moreover, Prasanth et al.15 suggested that the feather-edge offered advantages in tooth reduction, veneer preparation and cementation.\n\nOn the other hand, the benefits of the palatal chamfer margin can be explained both mechanically and adhesively. The mechanical advantage of the palatal extension of ceramic veneer is that it serves as a shear key and holds the veneer against the tooth. The adhesive advantage is the increased surface area for adhesive interface bonding on the palatal aspect16. Hence, it is critical to understand whether different preparation designs can affect longevity of ceramic veneers.\n\nIn this research, 0.5–0.8 mm labial reduction was performed to assure that the reduction is restricted to enamel, which increases the bonding and the strength without over-contouring17,18.\n\nThe materials used in this study were lithium disilicate and Cerasmart. IPS e.max lithium disilicate glass ceramic has a distinctive mixture of strength and optical characteristics and is available as pressable ingots or machinable blocks. The higher flexural strength and fracture toughness of IPS e.max press over e.max CAD made it the material of choice in our study19. The Cerasmart hybrid ceramic was selected as it is less brittle, more flexible and has better stress-absorbing characteristics than standard ceramics20.\n\nIn vitro simulation of oral conditions allows better evaluation of performance of dental materials. Thus, thermal cycling was used to simulate oral cavity thermal modifications that can introduce stresses to the bonded interfaces21.\n\nOur fracture resistance results for both IPS e.max lithium disilicate and Cerasmart hybrid materials showed non-significant differences, indicating that Cerasmart is a comparable material to IPS e.max and may represent alternative material for laminate veneers. This result could be related to the favorable association of low flexural modulus and proper flexural strength of the hybrid material, which improves the capacity to resist loading by more elastic deformation before failure22. By contrast, ceramic materials showed comparatively elevated flexural strengths and flexural modules, which reduced the capacity to undergo deformation to absorb the stress of enhanced loading23. Another crucial aspect that illustrates this point is the behavioral synergy between the hybrid and adhesive structure with comparable compositions and elevated bonding ability18.\n\nRegarding the fracture mode, it could be linked to the materials’ flexural strength and modulus of elasticity. Whenever the veneer’s flexural strength cannot offer tooth protection, its fracture will occur to prevent the delivery of the applied force to the tooth24. Additionally, the low modulus of elasticity correlates to increased deformation under load18. Accordingly, Cerasmart veneers were more likely to absorb the stresses which is considered a benefit to endure intra-oral forces and protect the underlying tooth structure against fracture. The e.max veneers showed only teeth fractures (cervical/root) following stresses.\n\nCastelnuovo et al.12 reported the presence of coronal, cervical and root fractures of teeth could be restored with leucite glass ceramic veneers. This is because the enamel not only generates an extremely predictable and stable bond, but also gives the tooth a stiffness that seems appropriate to replicate the tooth’s initial stiffness25.\n\nThe findings of this study showed that the feather-edge group had a statistically non-significant reduced mean value of fracture resistance than the wrap-around group in terms of preparation design, regardless of the material.\n\nThis was in agreement with the results of Highton et al.26, showing that the incisal wrap-around design reduces the stresses in laminate veneers by distributing the occlusal force to a wider region. Additionally, De Andrade et al.27 found that incisal wrap-around design were three times more resistant to the axial forces than feather-edge design. Moreover, Duzyol et al.28 found that incisal overlap design had the highest values of fracture resistance and justified this by their efficient ability to distribute the applied forces on the teeth.\n\nConversely, Hui et al.29 showed that the overlap design will transmit maximum stress on the veneer and increase the risk of cohesive fracture. In a systematic review by Albanesi et al.30, ceramic laminate veneers generally had high survival rates regardless of the preparation designs including incisal edge or not.\n\nIn the current study, the mean value of the force of fracture was greater than average chewing forces (20–160 N) recorded in the anterior teeth9. Consequently, our findings presented are clinically important.\n\nFinally, our ultimate goal in prosthetic dentistry is to provide our patients with the most conservative and satisfactory results. Therefore, careful selection of the most suitable restorative material and tooth preparation design for each clinical case needs to be done.\n\nThe test used in this study is considered a limitation, as the specimens were subjected to static rather than cyclic loading.\n\n\nConclusions\n\n1. The material used in this study for fabrication of the laminate veneers restorations has no crucial effect on its performance with regard to fracture resistance. Thus, the Cerasmart hybrid material could be considered a valid alternative to IPS e.max material.\n\n2. The fracture resistance of laminate veneers is not influenced by different preparation designs (feather-edge and wrap-around).\n\n3. Cerasmart veneers are more likely to absorb stresses and protect the underlying tooth structure than e.max veneers.\n\n\nData availability\n\nOpen Science Framework: Evaluation of Fracture Resistance of Cerasmart and Lithium Disilicate Ceramic Veneers with Different Incisal Preparation Designs. (In vitro Study). https://doi.org/10.17605/OSF.IO/83N9631.\n\nThis project contains the following underlying data:\n\n• Fracture resistance (All groups).xlsx (containing the fracture resistance load in N).\n\n• Fracture resistance (Paired groups).xlsx (containing comparisons of the indicated groups).\n\nData are available under the terms of the Creative Commons Zero “No rights reserved” data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nWe thank Omnia Nabil [MSc, Phd, Fixed Prosthodontics Department, Faculty of Dentistry, Cairo University] for her efforts, valuable guidance and support for completion of this project.\n\n\nReferences\n\nSadighpour L, Geramipanah F, Allahyari S, et al.: In vitro evaluation of the fracture resistance and microleakage of porcelain laminate veneers bonded to teeth with composite fillings after cyclic loading. J Adv Prosthodont. 2014; 6(4): 278–284. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGuess PC, Vagkopoulou T, Zhang Y, et al.: Marginal and internal fit of heat pressed versus CAD/CAM fabricated all-ceramic onlays after exposure to thermo-mechanical fatigue. J Dent. 2014; 42(2): 199–209. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDirxen C, Blunck U, Preissner S: Clinical performance of a new biomimetic double network material. Open Dent J. 2013; 7: 118–122. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAkoğlu B, Gemalmaz D: Fracture resistance of ceramic veneers with different preparation designs. J Prosthodont. 2011; 20(5): 380–384. PubMed Abstract \| Publisher Full Text\n\nAlghazzawi TF, Lemons J, Liu PR, et al.: The failure load of CAD/CAM generated zirconia and glass-ceramic laminate veneers with different preparation designs. J Prosthet Dent. 2012; 108(6): 386–393. PubMed Abstract \| Publisher Full Text\n\nJankar AS, Kale Y, Kangane S, et al.: Comparative evaluation of fracture resistance of Ceramic Veneer with three different incisal design preparations - An In-vitro Study. J Int Oral Heal. 2014; 6(1): 48–54. PubMed Abstract \| Free Full Text\n\nArora A, Upadhyaya V, Arora SJ, et al.: Evaluation of fracture resistance of ceramic veneers with different preparation designs and loading conditions: An in vitro study. J Indian Prosthodont Soc. 2017; 17(4): 325–31. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPini NP, Aguiar FH, Lima DA, et al.: Advances in dental veneers: materials, applications, and techniques. Clin Cosmet Investig Dent. 2012; 4: 9–16. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSoulieman H, Saker R: Fracture resistance of feldspathic veneers with different preparation designs in vitro. Int J Dent Health Sci. 2017; 4(2): 247–255. Reference Source\n\nSoares CJ, Pizi EC, Fonseca RB, et al.: Influence of root embedment material and periodontal ligament simulation on fracture resistance tests. Braz Oral Res. 2005; 19(1): 11–6. PubMed Abstract \| Publisher Full Text\n\nSterzenbach G, Kalberlah S, Beuer F, et al.: In-vitro simulation of tooth mobility for static and dynamic load tests: a pilot study. Acta Odontol Scand. 2011; 69(5): 316–8. PubMed Abstract \| Publisher Full Text\n\nCastelnuovo J, Tjan AH, Phillips K, et al.: Fracture load and mode of failure of ceramic veneers with different preparations. J Prosthet Dent. 2000; 83(2): 171–180. PubMed Abstract \| Publisher Full Text\n\nMeijering AC, Creugers NH, Roeters FJ, et al.: Survival of three types of veneer restorations in a clinical trial: a 2.5-year interim evaluation. J Dent. 1998; 26(7): 563–568. PubMed Abstract \| Publisher Full Text\n\nFriedman MJ: A 15-year review of porcelain veneer failure--a clinician's observations. Compend Contin Educ Dent. 1998; 19(6): 625–628, 630, 632 passim; quiz 638. PubMed Abstract\n\nPrasanth V, Harshkumar K, Lylajam S, et al.: Relation between fracture load and tooth preparation of ceramic veneers - Anin vitro study. Health Sci. 2013; 2: 1–11.\n\nSchmidt KK, Chiayabutr Y, Phillips KM, et al.: Influence of preparation design and existing condition of tooth structure on load to failure of ceramic laminate veneers. J Prosthet Dent. 2011; 105(6): 374–382. PubMed Abstract \| Publisher Full Text\n\nBergoli CD, Meira JB, Valandro LF, et al.: Survival rate, load to fracture, and finite element analysis of incisors and canines restored with ceramic veneers having varied preparation design. Oper Dent. 2014; 39(5): 530–540. PubMed Abstract \| Publisher Full Text\n\nKhaliq AGA, Al-Rawi II: Fracture strength of laminate veneers using different restorative materials and techniques (A comparative in vitro study). J Baghdad Coll Dent. 2014; 26(4): 1–8. Publisher Full Text\n\nZarone F, Ferrari M, Mangano FG, et al.: \"Digitally Oriented Materials\": Focus on Lithium Disilicate Ceramics. Int J Dent. 2016; 2016: 9840594. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAl-shibri S, Elguindy J: Fracture Resistance of Endodontically Treated Teeth Restored With Lithium Disilicate Crowns Retained With Fiber Posts Compared To Lithium Disilicate and Cerasmart Endocrowns: In Vitro Study. Dentistry. 2017; 7: 12. Publisher Full Text\n\nMazzitelli C, Monticelli F, Toledano M, et al.: Effect of thermal cycling on the bond strength of self-adhesive cements to fiber posts. Clin Oral Investig. 2012; 16(3): 909–915. PubMed Abstract \| Publisher Full Text\n\nSwain MV, Coldea A, Bilkhair A, et al.: Interpenetrating network ceramic-resin composite dental restorative materials. Dent Mater. 2016; 32(1): 34–42. PubMed Abstract \| Publisher Full Text\n\nAwada A, Nathanson D: Mechanical properties of resin-ceramic CAD/CAM restorative materials Presented at the American Association of Dental Research/Canadian Association of Dental Research Annual Meeting, Charlotte, NC, March 2014. J Prosthet Dent. 2015; 114(4): 587–593. PubMed Abstract \| Publisher Full Text\n\nTomer A, Raina A, Bin Ayub F, et al.: Fracture strength of composite veneers using different restorative materials: A comparative in vitro study. IJADS. 2017; 3(4): 465–468. Reference Source\n\nLin TM, Liu PR, Ramp LC, et al.: Fracture resistance and marginal discrepancy of porcelain laminate veneers influenced by preparation design and restorative material in vitro. J Dent. 2012; 40(3): 202–209. PubMed Abstract \| Publisher Full Text\n\nHighton R, Caputo AA, Mátyás J: A photoelastic study of stresses on porcelain laminate preparations. J Prosthet Dent. 1987; 58(2): 157–161. PubMed Abstract \| Publisher Full Text\n\nDe Andrade OS, Hirata R, Celestrino M, et al.: Ultimate ceramic veneer: a laboratory-guided preparation technique for minimally invasive laminate veneers. J Calif Dent Assoc. 2012; 40(6): 489–494. PubMed Abstract \| Publisher Full Text\n\nDuzyol M, Duzyol E, Seven N: Fracture Resistance of Laminate Veneers Made with Different Cutting and Preparation Techniques. Int J Dent Sci Res. 2016; 4(3): 42–48. Reference Source\n\nHui KK, Williams B, Davis EH, et al.: A comparative assessment of the strengths of porcelain veneers for incisor teeth dependent on their design characteristics. Br Dent J. 1991; 171(2): 51–55. PubMed Abstract \| Publisher Full Text\n\nAlbanesi RB, Pigozzo MN, Sesma N, et al.: Incisal coverage or not in ceramic laminate veneers: A systematic review and meta-analysis. J Dent. 2016; 52: 1–7. PubMed Abstract \| Publisher Full Text\n\nMohammed B: Evaluation of Fracture Resistance of Cerasmart and Lithium Disilicate Ceramic Veneers with Different Incisal Preparation Designs. (In vitro Study). 2019. http://www.doi.org/10.17605/OSF.IO/83N96" }	[ { "id": "67491", "date": "27 Jul 2020", "name": "Khalid M. Abdelaziz", "expertise": [ "Reviewer Expertise Restorative Dentistry & Dental materials" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article aimed to evaluate the fracture resistance of two different laminate veneer restorative materials. The influence of the preparation design was also considered. The selected materials offer different characteristics based on their structural natures and compositions. Therefore, comparing the performance of these materials in situations mimicking that of the normal intraoral environment is of great importance and would guide the practicing dentist to select the most clinically beneficial material.\n\nGeneral comment:\nThe whole article should be linguistically revised. Rephrasing of some sentences is highly recommended.\n\nTitle:\n\nAfter going through this article, I do suggest some changes to the study title “Fracture Resistance of Teeth Restored with Hybrid and Glass-based Ceramic Veneers”. The new title, according to my vision, would be beneficial for the work and for properly leading the readers.\n\nAbstract:\n\nThe aim of the study should be restructured according to the new title. The conclusion should also be redrawn according to the new aim.\n\nIntroduction:\nShould be revised and modified wherever possible according to the suggested aim. Highlights on the actually utilized composite material (as shown at the beginning of the abstract section) should be considered. Referring to previous literature (if any) that showed support/conflict of using these materials would be helpful. The aim should be modified too.\n\nMaterials and methods:\nA table showing the detailed description, composition and manufacturer of the used materials must be added. This will help the reader to easily discover the differences between both categories of the restorative materials.\n\nTesting procedure: The recorded test results were greatly influenced by the bulk of tooth structure. Referring to the testing mechanism, the site of load application would flex the tooth at the cervical third. In the tooth crown the cervical third is having the most bulk of the tooth structure and accordingly shows minimal flexing. Most of the applied force would affect the cervical constriction (at the cervical line) causing fracture of the tooth itself. It is so important for the authors to mention when they exactly recorded the fracture load (must be at the first-heard cracking/clicking sound) simply because if the fracture starts within the tooth at the cervical line, continuous loading could lead to fracture or debonding of the veneer leading to false results.\nTherefore, the influence of tooth flexing on the bonded restoration was; accordingly, too minimal to induce fracture unless defective cementation was there to initiate cracking (this can explain the high values of the SD and the insignificant differences between the recorded results). Therefore, in my opinion, the performed test was evaluating the fracture resistance of the restored teeth rather than the fracture resistance of the restorations themselves. To test the fracture resistance of the veneering materials, 3-point flexure test is recommended to test bar specimens obtained from the restored teeth.\n\nResults:\n\nDuplicate presentation of the results (Tables and graphs) is not recommended.\n\nDiscussion and Conclusion:\nShould be adjusted according to the aforementioned suggestions.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "78234", "date": "03 Mar 2021", "name": "Uwe Blunck", "expertise": [ "Reviewer Expertise Adhesive Dentistry: adhesion to enamel", "dentin", "ceremic", "composite." ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn general: The authors are using incorrect terms for describing the tested material: the tested CERASMART is not a hybrid ceramic! Today, a large number of tooth-colored materials for CAD/CAM milling machines are available. These include silicate ceramics (i.e. feldspar ceramics, leucite-reinforced glass ceramics, lithium disilicate-based ceramics, and zirconia-reinforced lithium silicate ceramics), oxide ceramics (i.e. zirconium dioxide ceramics), and a new category of hybrid materials (organic resin matrix highly filled with porcelain particles) recently classified as resin matrix ceramics: these can be further subdivided in ceramics with a “resin-infiltrated matrix”, referring to the so-called polymer-infiltrated ceramic network, and “resin-based composites”, referring to industrial polymerized resin-based composites produced under high-pressure and in a high-temperature environment. The composition of Cerasmart (GC Corp, Tokyo, Japan): Bis-MEPP, UDMA, DMA, 71 wt% of Silica (20 nm) and barium glass (300 nm).\nFrom the literature: CERASMART® blocks (GC Dental products Europe, Belgium) are composite resin nanoceramic blocks that consist of a polymeric matrix reinforced by ceramic nanohybrid ﬁllers. Goujat et al. (20181).\nThe company itself refers to a publication (Lauvahutanon et al., 20142) which defines Cerasmart as a CAD/CAM composite resin block.\nThis need to be corrected within the entire paper.\nThe manuscript also needs the following corrections: Abstract: Change to: “All samples were subjected to thermocycling between 5°C and 55°C in a water bath for a total of 1750 cycle with 30 seconds dwell time at each bath.” Change to: “The fracture resistance of laminate veneers is not influenced by different types of preparation designs.”\n\nTitle: This reviewer also supports the suggestion of the first reviewer for the title “Fracture Resistance of Teeth Restored with Hybrid and Glass-based Ceramic Veneers”. (see below).\n\nMethods: The term “wrap-around preparation” is not common in the scientific literature. In Figure 2 it looks like a butt joint preparation and it might be helpful for the reader to get more specific details of this preparation design.\nThe description for the loading process “The load was applied using a metallic rod with flat tip (3.8 mm diameter) attached to the upper movable compartment of testing machine traveling at cross-head speed of 1 mm/min with tin foil sheet in-between to achieve homogenous stress distribution and minimization of the transmission of local force peaks” clearly shows that the mechanical stress is transferred to the tooth and not to the veneer itself. Therefore, this study evaluated the measurement of the stability of the incisors and not of the different veneers.\nConditioning of the veneers: The authors describe: “the fitting surfaces of the veneers were etched with 9.5% hydrofluoric acid gel IPS Ceramic Refill (Ivoclar Vivadent AG, Principality of Liechtenstein)” There is one hydrofluoric acid on the market (Porcelain Etch, Ultradent) with the concentration of 9.5 % HF, however, IPS Ceramic Etch from Vivadent has only a concentration of 4.5 % HF!\nFracture resistance test: If the load is - as described before - applied to the “lingual surface of the tooth” the last sentence of this paragraph should be changed to: “The load required to either fracture of the tooth or debonding of the veneer was recorded in Newtons.” Otherwise, the end of the loading should have resulted in tooth fracture, different from Figure 10.\n\nResults: The sentence “E.max wrap-around group recorded statistically non-significant (P>0.05) highest fracture resistance mean value (422.1 N) followed by the Cera wrap-around group (317.23 N), and then E.max feather-edge group (289.6 N), while Cera featheredge group recorded the lowest fracture resistance mean value (259.3 N)” is misleading. Whenever a statistical evaluation finds no significant difference between the results of two groups, the two groups are considered equal. However, the sentence suggests that there are differences, which is from a statistical point of view not the case, since the differences are - as seen from the statistics - just by incidence.\n\nDiscussion: It would be helpful for the readers to mention in the discussion why the Cerasmart CAD-CAM composite resin blocks can be conditioned by hydrofluoric acid. Usually it is recommended to pretreat CAD-CAM composite resin blocks by sandblasting with Al2O3 (Lise et al., 20173).\nHowever, Cekic-Nagas et al. reported that HA had no effect on the bond strength in CER, ENA and LAV specimens. The main compositions of these materials are silica and silicate compounds, and HA partially dissolved those silicas (Cekic-Nagas et al., 20164).\n\nManufacturer’s instruction recommends both approaches: CEMENTATION: With Sandblasting technique: 1) Sandblasting with 25-50μm alumina 0.15MPa/1.5bar is recommended. 2) Blow the restoration with an oil-free air syringe or clean with anultrasonic cleaner and dry. 3) Clean with alcohol to remove oil residue. 4) Treat the surface with a silane coupling agent such as CERAMIC PRIMER II or G-Multi PRIMER. 5) Cement with an adhesive resin cement such as G-CEM LinkForce.\nWith hydrofluoric acid etching: 1) Treat with hydrofluoric acid (5%) for 60 seconds. 2) Wash with water spray or an ultrasonic cleaner and dry. 3) Treat the surface with a silane coupling agent such as CERAMIC PRIMER II or G-Multi PRIMER. 4) Cement with an adhesive resin cement such as G-CEM LinkForce.\n\nThe authors write: “The findings of this study showed that the feather-edge group had a statistically non-significant reduced mean value of fracture resistance than the wrap-around group in terms of preparation design, regardless of the material.” Whenever a statistical evaluation finds no significant difference between the results of two groups, the two groups are considered equal. Therefore, the sentence should be revised.\n\nConclusion: The authors write: “Cerasmart veneers are more likely to absorb stresses and protect the underlying tooth structure than e.max veneers.” This statement by the authors is based on the results that in the two Cerasmart groups 5 and 7 teeth respectively were intact after the loading. However, just the same numbers were found for debonded veneers. It should have been at least discussed in the paper whether the reason could also have been the less effective bonding of the luting resin to the CAD-CAM composite resin, which also might result in a loss of retention of the veneers.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [] } ]	1	https://f1000research.com/articles/8-1491
https://f1000research.com/articles/8-1490/v1	23 Aug 19	{ "type": "Software Tool Article", "title": "A Sequence Distance Graph framework for genome assembly and analysis", "authors": [ "Luis Yanes", "Gonzalo Garcia Accinelli", "Jonathan Wright", "Ben J. Ward", "Bernardo J. Clavijo", "Luis Yanes", "Gonzalo Garcia Accinelli", "Jonathan Wright", "Ben J. Ward" ], "abstract": "The Sequence Distance Graph (SDG) framework works with genome assembly graphs and raw data from paired, linked and long reads. It includes a simple deBruijn graph module, and can import graphs using the graphical fragment assembly (GFA) format. It also maps raw reads onto graphs, and provides a Python application programming interface (API) to navigate the graph, access the mapped and raw data and perform interactive or scripted analyses. Its complete workspace can be dumped to and loaded from disk, decoupling mapping from analysis and supporting multi-stage pipelines. We present the design and implementation of the framework, and example analyses scaffolding a short read graph with long reads, and navigating paths in a heterozygous graph for a simulated parent-offspring trio dataset. SDG is freely available under the MIT license at https://github.com/bioinfologics/sdg", "keywords": [ "Genome graph", "genome assembly" ], "content": "Introduction\n\nSequence graphs are the core representation of genome assemblers1–3 Their use has increased lately thanks to the graphical fragment assembly (GFA) format for graph exchange4, tools to work with genome variation graphs5, and sequence to graph mappers6–10 But a lack of inter operation between graph-based tools, and limited tools for downstream graph-based analysis, contribute to a perceived complexity which maintains linear sequences as the typical unit of exchange. This flattening of graph representations within pipelines with multiple steps, that use different types of sequencing in an iterative fashion, produces ever-longer linear genome sequences through an information loss process. As a result, genome assembly projects are prone to error propagation and difficult to reproduce and control. These problems can be addressed developing graph-based frameworks to integrate the analysis of hybrid datasets.\n\nThe Sequence Distance Graph (SDG) framework implements a SequenceDistanceGraph representation that defines sequences in nodes and their adjacency in links, and an associated Workspace containing raw data and mappings. This provides an integrated working environment to use multiple sources of information to navigate and analyse genome graphs. Datastores allow random access to short, linked, and long read sequences on disk. A mapper on each datastore contains methods to map the reads to the graph and access the mapping data. KmerCounters provide functions to compute k-mer coverage over the graph from sequencing data, enabling coverage analyses. Additional DistanceGraphs, typically representing longer-range information and different linkage levels, define alternative topologies over the SequenceDistanceGraph nodes. Finally, a NodeView abstraction provides a proxy to a node, with methods to navigate the graph and access its mapped data. This comprehensive framework can be used to explore genome graphs interactively or to create processing methods for assembly or downstream analysis.\n\nHere we describe the SDG implementation and basic tools, providing examples of use cases that highlight its analytic flexibility. First, we show how to create a hybrid assembly by integration of long reads linkage into a short-read graph. Then we analyse a simulated parent-child trio and show how the coverage of the parent datasets can be used to navigate the graph topology. These are only two of the multiple ways integrating data and genome graphs can be used to perform simple but powerful analyses.\n\n\nMethods\n\nThe C++ core library implements SDG’s data structures and methods for WorkSpaces, graphs, datastores and mappers. Its main goal is to provide a straightforward interface to project information from raw datasets onto graphs, and enable easy access and analysis of the graph-data combination. It uses OpenMP for parallel processing, and SWIG 4.0 to export a Python API to enable interactive data analysis.\n\nThe SequenceDistanceGraph class contains a vector of nodes defining DNA sequences, and a vector of links. Every node has a positive and a negative end, and links are defined between these node ends. Links with positive distances represent gaps between linked sequences and negative distances represent overlaps. This representation, shown in Figure 1, is similar to those presented in 2,11 but unifies the concept of overlap and gap. Paths can be defined as list of nodes, with the sign of the first end in the walk. Graphs can be read and written to GFA and GFA2 files.\n\nSequences appear in only one direction and their reverse complement can be obtained by traversing the node in opposite direction, from - to +. The two largest possible paths are [1, 2, 4, 5] and [1, -3, 4, 5], and their reverse complements [-5, -4, -2, -1] and [-5, -4, 3, -1] respectively.\n\nThe DistanceGraph class contains a set of links over the nodes of a SequenceDistanceGraph object. It is used to represent alternative sources of linkage information, such as longer range linkage produced by mapped reads for scaffolding.\n\nThe WorkSpace contains a single SequenceDistanceGraph, multiple DistanceGraphs, datastores and mappers, and its structure in memory represents the status of the SDG framework. It can be dumped and loaded from disk, providing persistence and checkpoints between different steps on SDG-based pipelines. Raw reads and k-mer counts are kept in separate files, pointed from the WorkSpace, to avoid duplication when using multiple WorkSpaces around the same dataset.\n\nThe DataStores and Mappers provide access and management to raw data and its mapping on the graph. Datastores do not load read data into memory, but rather provide random access to the on-disk data. The PairedRedMapper and LinkedReadMapper classes use a unique k-mer index to map reads to single nodes, with single reads mapping to multiple nodes not being mapped3,12 The LongReadMapper class generates multiple mappings from each read to nodes, using a short non-unique k-mer index (k=15 by default)13,14 Long read mapping filtering is left to later stages of the processing.\n\nThe KmerCounters creates an index with all the k-mers at a given k up to k=31 and counts occurrences of these k-mers on the graph, allowing then to count occurrences in datastores or fastq files. These counts, persisted in the KmerCounter with a name, can be then accessed to perform k-mer coverage analyses. Projections of raw k-mer coverage in the reads and the assembly over a particular sequence for a node or path, similar to those produce by the \"sect\" tool of K-mer Analysis Toolkit (KAT)15 are valuable for content analysis. Spectra analysis of these frequencies can provide further insight into genome composition and representation on the assembly.\n\nTwo processing classes, LinkageUntangler and LinkageMaker, work with alternative linkage configurations. The LinkageMaker is used to condense information via one of its make_linkage* methods, from evidence in the WorkSpace into links in a DistanceGraph. The LinkageUntangler class works on a DistanceGraph to simplify, condense and/or linearise its linkage. In the second use case below it can be seen how a combination of LinkageMaker and LinkageUntangler can be used for scaffolding with long reads.\n\nFinally, the NodeView class, and its associated LinkViews, provide a single-entry point for node-centric analyses. A NodeView from either a DistanceGraph or SequenceDistanceGraph is a wrapper containing a pointer to the graph and a node id, and will provide access to its nodes’ previous and next linked nodes, mapped reads, or k-mer coverage. A user with good understanding of the NodeView class should be able to access most information in the WorkSpace through it, making it the default choice for analysing the graph.\n\nRequirements and installation. SDG can be run on Linux and MacOS, and requires enough RAM to hold the WorkSpace completely in memory, which will depend on the dataset. Space to hold the uncompressed sequences on the datastores on disk will also be required.\n\nSDG can be installed via pre-compiled binaries from https://github.com/bioinfologics/sdg/releases. The binaries have been built using Python3 and GCC version 6 from the Ubuntu package manager for the Linux version. The MacOS version dependencies were obtained using Homebrew (Python3, GCC-6 and SWIG). SDG can be compiled using CMake, Python3, SWIG version 4 and GCC version 6 onwards. Detailed instructions can be found at https://bioinfologics.github.io/sdg/sdg/README.html#installation.\n\nTypical workflow. Working with SDG typically involves two different stages: creating a WorkSpace with the data and mappings, and analysing this WorkSpace. SDG includes command line tools to create DataStores, KmerCounts, and WorkSpaces, and map reads within a WorkSpace.\n\nsdg-datastore: creates a Datastore from raw reads and can process paired, 10x or long reads. An output prefix is specified as a parameter and a <prefix>.prseq, <prefix>.lrseq or <prefix>.loseq file is generated.\n\nsdg-kmercounter: creates a KmerCounter indexing a graph from a WorkSpace or GFA, or works with an already generated one. A count can be added directly from raw reads or from a datastore. The KmerCounter is persisted on file with extension ’sdgkc’.\n\nsdg-workspace: creates a WorkSpace from a base graph or works with an already generated one. Datastores and KmerCounters can be added. The WorkSpace is persisted on file with extension ’sdgws’.\n\nsdg-dbg: creates a WorkSpace from a PairedReadDatastore by building a deBruijn graph and using this as the base graph. Counts for the k-mers from the graph and raw reads are added too.\n\nsdg-mapper: maps reads within a WorkSpace. An updated WorkSpace is produced and dumped to the specified prefix.\n\nWorkSpaces can also be instantiated with an empty graph, and the graph populated through the add_node and add_link methods. The following example on a python session shows how the simple graph from Figure 1 can be created from scratch, navigated through a NodeView instance and sequence from its paths extracted.\n\n\n\nTypically, as shown in Figure 2, the API is used to explore a larger WorkSpace, with the methods accessing both in-memory and on-disk data, and modifying the status of the WorkSpace.\n\nThe WorkSpace holds the information for a project and contains the graphs, the mappers and k-mer counts. From Python, a previously saved WorkSpace is loaded from disk (1). The NodeView object is centred on a specific node and can be used to access node characteristics (ie. size and sequence), graph topology from the perspective of the node you are on (i.e. neighbours in both directions (2)) and can also retrieve information projected onto the selected node (ie. mappings (3) and k-mer coverage (4)). Operations such as adding a KmerCounter to the WorkSpace and adding a count (5) can be performed, and the WorkSpace can be saved back to disk (6). Once loaded, the bulk of the WorkSpace is held in memory for fast access with the raw read data from the DataStores remaining on disk accessible through random access.\n\n\nExample use cases\n\nTo illustrate the use of SDG, we have reproduced a short version of two examples from http://bioinfologics. github.io/sdg_examples.\n\nAll paired end datasets are available on https://zenodo.org/record/3363871#.XUwyVy2ZN2416, and the PacBio reads are from NCBI accession PRJNA19443717 For simplicity, we have also made the datasets available on https://opendata.earlham.ac.uk/opendata/data/sdg_datasets/ as ready-to-use ’fastq.gz’ files.\n\nThis example is based on an E. coli dataset combining PacBio reads from 17 and Illumina Miseq 2x300bp reads subsampled from a test run. It uses the long reads to scaffold a short read based graph produced by sdg-dbg. Graphs are dumped to GFA files at different stages, and visualised using Bandage v0.8.118\n\nFirst, we use the command line tools to create datastores for both long and short reads and an initial WorkSpace containing a DBG assembly:\n\n\n\nFrom this point on, we use the python SDG library. First, we load the workspace, add a long read datastore and map its reads using a k=11 index.\n\n\n\nThe graph, as shown in Figure 3A contains multiple unresolved repeats.\n\nLinkage at different stages of the long read scaffolding example, visualised using Bandage: A) SequenceDistanceGraph generated by sdg-dbg from short reads, B) DistanceGraph generated after using make_nextselected_linkage on the long read data, linking all nodes of 1100bp and more, C) DistanceGraph after eliminating all nodes with multiple connections (repeats).\n\nWe can use the LinkageMaker to create linkage using the long reads datastore. We do this by selecting the nodes between which to analyse possible linkage, in this case all nodes of 1100bp or more, and then calling the make_longreads_multilinkage method, with alignment filtering parameters of 1000bp and 10% id.\n\n\n\nThis multi-linkage can be collapsed using the LinkageUntangler. The make_nextselected_linkage method links every selected node to its closest selected neighbours on each direction, aggregating the distances via a simple median calculation:\n\n\n\nThe new graph we dumped, as shown in Figure 3B, has disconnected the repeats and introduced long read linkage which skips over them, but it is still not fully solved. We can improve this further by getting rid of repetitive nodes that will be connected to multiple neighbours, as each of them belongs in more than one place. We do that by just turning these nodes’ selection off in the LinkageUntangler, which will then skip them in the solution.\n\n\n\nThe last graph is now a circle, with all the repeats disconnected from any linkage.\n\nWe created a simulation of a trio dataset for this example using the synthetic genome creation and sequencing package Pseudoseq.jl v0.1.019 Chromosomes 4 and 5 of the reference genome of the yeast strain S288C were used as templates to create a diploid, genome for each parent with 1% heterozygous sites. Each homologous pair of chromosomes was crossed over and recombined and the child inherited one chromosome from the first parent at random, and one chromosome from the second parent at random. Simulated paired end reads were generated for each genome, using an average fragment length of 700bp and a read length of 250bp, and an expected coverage of 70x with error rate was set to 0.1%.\n\nFirst we used the command line tools to create a graph from the child reads using sdg-dbg, and add k-mer counts for both parents into the datastore.\n\n\n\nWe now open the WorkSpace and use the NodeView::parallels method to look for the largest bubble structure in the graph, which should be formed by two parallel nodes with haplotypes coming from each parent.\n\n\n\nSince each side should be a haplotype from a different parent, we should see a loss of k-mer coverage on the parent that didn’t contribute that haplotype. To check this, we create a plotting function to plot the output from the NodeView::kmer_coverage method.\n\n\n\nThe plots, shown in Figure 4, reflect how Node 4775 contains content inherited from parent 2 and its parallel node 11414 contains content inherited from parent 1. We can create a function to extend these parent-specific regions by walking forward and backward as long as only one link takes us to a node that is fully covered by the content of the parent we are following.\n\nCoverage drops to 0 on the opposite parent for k-mers that are unique to a parent.\n\n\n\nAfter using this function, path1 contains 49 nodes yielding 8672bp of sequence inherited from parent 1, and path2 contains 139 nodes yielding 26351bp of sequence inherited from parent 2. It is important to note that the difference in node count and sequence length arises because the extension function is haplotype-specific and its results depend in the topology of each haplotype graph.\n\n\nSummary\n\nThe Sequence Distance Graph framework provides a unified workspace for different sequencing technologies using the genome graph as the basis of integration. It enables analyses across the graph topology, the raw data and its projections to the graph. We have shown how the NodeView class can be used through the Python API to produce interactive analyses that are both powerful and easy to follow. We expect this will be a useful codebase for all levels of users, not only for the construction of graph-based analysis but also for their teaching and dissemination.\n\n\nData availability\n\nThe PacBio, E. coli reads are deposited on NCBI accession PRJNA194437 from Koren et al.17\n\nE. coli K12 Re-sequencing with PacBio RS and 454: Accession number PRJNA194437, https://identifiers.org/ncbi/bioproject:PRJNA194437\n\nThe datasets used in the examples are available from: https://opendata.earlham.ac.uk/opendata/data/sdg_datasets/ and archived in Zenodo Zenodo: SDG Paper Datasets. http://doi.org/10.5281/zenodo.336387116\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).\n\n\nSoftware availability\n\nSoftware documentation: https://bioinfologics.github.io/sdg\n\nSource code available from: http://github.com/bioinfologics/sdg\n\nArchieved source code at time of publication: https://zenodo.org/record/3363165#.XUw1yy2ZN2520\n\nLicense: MIT License", "appendix": "Grant information\n\nThis work was strategically funded by the BBSRC Core Strategic Programme Grant [BBS/E/T/000PR9818] Work by GGA and BJC was also partially funded by the BBSRC grant \"OctoSeq: Sequencing the octoploid strawberry\" [BB/N009819/1].\n\n\nAcknowledgements\n\nWe would like to thank Richard Harrison for helpful discussions about SDG’s results and continued support through the OctoSeq project. We thank James Cuff for input about design principles and continuous encouragement. We thank Kat Hodgkinson for her feedback and patience as an early user of the rough alpha version of SDG. We thank Camilla Ryan for enduring and joining never-ending discussions about graph representations and the design of the framework.\n\n\nReferences\n\nPevzner PA, Tang H, Waterman MS: An Eulerian path approach to DNA fragment assembly. Proc Natl Acad Sci U S A. 2001; 98(17): 9748–9753. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMedvedev P, Brudno M: Maximum likelihood genome assembly. J Comput Biol. 2009; 16(8): 1101–1116. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nButler J, MacCallum I, Kleber M, et al.: ALLPATHS: de novo assembly of whole-genome shotgun microreads. Genome Res. 2008; 18(5): 810–820. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJackman SD, Myers EW, Gonella G: The GFA Specification. Reference Source\n\nGarrison E, Sirén J, Novak AM, et al.: Variation graph toolkit improves read mapping by representing genetic variation in the reference. Nat Biotechnol. 2018; 36(9): 875–879. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRautiainen M, Mäkinen V, Marschall T: Bit-parallel sequence-to-graph alignment. bioRxiv. 2018; 323063. Publisher Full Text\n\nSirén J, Garrison JE, Novak AM, et al.: Haplotype-aware graph indexes. bioRxiv. 2019. Publisher Full Text\n\nNovak AM, Garrison E, Paten B: A graph extension of the positional Burrows-Wheeler transform and its applications. Algorithms Mol Biol. 2017; 12(1): 18. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJain C, Dilthey A, Misra S, et al.: Accelerating Sequence Alignment to Graphs. bioRxiv. 2019. Publisher Full Text\n\nLimasset A, Flot JF, Peterlongo P: Toward perfect reads: self-correction of short reads via mapping on de Bruijn graphs. Bioinformatics. 2019; pii: btz102. PubMed Abstract \| Publisher Full Text\n\nPaten B, Zerbino DR, Hickey G, et al.: A unifying model of genome evolution under parsimony. BMC Bioinformatics. 2014; 15(1): 206. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBatzoglou S, Jaffe DB, Stanley K, et al.: ARACHNE: a whole-genome shotgun assembler. Genome Res. 2002; 12(1): 177–189. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSović I, Šikić M, Wilm A, et al.: Fast and sensitive mapping of nanopore sequencing reads with GraphMap. Nat Commun. 2016; 7: 11307. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPavetić F, Katanić I, Matula G, et al.: Fast and simple algorithms for computing both LCSk and LCSk+. arXiv: 1705.07279 [cs], 2017. Reference Source\n\nMapleson D, Garcia Accinelli G, Kettleborough G, et al.: KAT: a K-mer analysis toolkit to quality control NGS datasets and genome assemblies. Bioinformatics. 2017; 33(4): 574–576. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYanes L, Garcia Accinelli G, Ward BJ, et al.: Sdg paper datasets. 2019. https://zenodo.org/record/3363871\n\nKoren S, Harhay GP, Smith TP, et al.: Reducing assembly complexity of microbial genomes with single-molecule sequencing. Genome Biol. 2013; 14(9): R101. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWick RR, Schultz MB, Zobel J, et al.: Bandage: interactive visualization of de novo genome assemblies. Bioinformatics. 2015; 31(20): 3350–3352. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWard BJ: bioinfologics/pseudoseq.jl: First release. 2019. Publisher Full Text\n\nYanes L, Garcia Accinelli G, Ward BJ, et al.: bioinfologics/sdg: Release candidate. 2019; 7. https://zenodo.org/record/3363165" }	[ { "id": "52959", "date": "24 Sep 2019", "name": "Erik Garrison", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors demonstrate a new toolchain and data model for working with sequence graphs. This method allows the user to dynamically interact with sequence graphs made in the process of assembly. They provide a number of examples of the use of the method as well as code snippets to demonstrate its functionality. The library is written in C++, but wrapped in python with SWIG, which should make it useful to many researchers for whom C++ is difficult to use.\n\nI find only one thing strange about the work. In the beginning, the authors indicate that there are not interoperable methods for working with sequence graphs and alignments to them, but they have in effect created another competing standard. Are there particular limitations with existing data models that they hope to address with the Sequence Distance Graph framework? How is their model different than the variation graph model, in which distances are provided by a collection of paths (or equivalently alignments) embedded within the sequence graph?\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] }, { "id": "55323", "date": "04 Nov 2019", "name": "Eric T. Dawson", "expertise": [ "Reviewer Expertise computational biology", "graph genomes", "structural variant calling", "bioinformatics", "cancer biology" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors describe a framework for constructing sequence graphs, aligning reads, manipulating graph structures, and extracting them into standard formats (e.g., GFA). This framework is available as both a set of command line tools and a python library which wraps much of the underlying functionality. Their implementation unifies the representation of gaps and overlaps as a single linkage type within the graph. This is the primary theoretical advance of the work. This work is scientifically sound but its description as written could benefit from some minor additions.\nThe software is freely available on GitHub and binary releases of the command line tools are provided. These are functional on a modern linux laptop and clear examples with data are provided. The paper includes the outputs of these examples as figures.\nThe python libraries rely on SWIG and are not included in the binaries. While not requisite for publication, providing the python libraries through pip, conda, or another package manager would increase the reach of the framework. This would match the authors' conclusion that the sdg package provides \"a useful codebase for all levels of users.\"\nThe examples provided are clear and scientifically relevant. The graph mapping and manipulation example using E. coli data (Figure 3) and the description of genotyping a simulated yeast trio (Figure 4) are both realistic.\nHowever, the authors should provide run times and machine details for these examples. Both are relatively fast as the datasets are small. There is no need for extensive benchmarking; a footnote for each example would address this adequately.\nA brief 1-2 sentence discussion of a larger scale example the authors have attempted should also be included.\nIn addition, the phrasing \"simulated parent-offspring trio\" in the abstract should be modified to make it clear that the data is from yeast. As it is written the phrasing implies the framework may work on human/animal-scale data, though no evidence of this has been provided in this version of the paper.\nLastly, a brief description of the similarities and differences between the sequence (distance) graph, the variation graph, and the de Bruijn graph from an assembler such as ABySS should be included in the introduction or provided by a reference. This description need not be longer than two to four sentences in length. This should highlight the similar representations of the graphs (e.g., sequences stored in nodes and linkages/paths described by edges) and the different amounts of information content within the graph types. This would strengthen the critical need for the software and is partially highlighted by the example in Figure 3.\nAs it stands the paper is deserving of indexing. These additions would further strengthen what is already an excellent tool description, I hope without adding too much additional work for the authors.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] }, { "id": "55755", "date": "24 Jan 2020", "name": "Max Alekseyev", "expertise": [ "Reviewer Expertise genome assembly", "comparative genomics" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors describe a software package aimed at construction, storage, and manipulation of sequence graphs. The software is publicly available on Github. While the core functionality is developed in C++, the package also provides Python wrapper library and a command-line tool. The authors outline the software design and provide pipelines and code examples for two different types of data.\n\nOverall, the paper is well-written and describes potentially useful software. At the same time, the paper lacks:\nComparison of (features of) the developed software and existing software such as VG toolkit based on variation graphs (reference 51);\n\nDiscussion of the software applicability (e.g., will it work on large or repeat-rich genomes?), or estimation of its running time/space complexity.\nSome minor comments:\nThe use of bold font is not explained. For example, Datastores first appears in the sentence “Datastores allow random access...” describing its features, but WHAT is Datastores?\n\nOrthogonal edge routing in Fig. 1 is somewhat confusing, why not make edges curved?\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-1490
https://f1000research.com/articles/8-1487/v1	23 Aug 19	{ "type": "Research Article", "title": "Antibacterial effect of the hydroalcoholic extract of Mauritia flexuosa leaves on gram-negative and gram-positive bacteria", "authors": [ "Ana Sandoval Vergara", "Marlon Farfán Córdova", "Marco Leoncio Salazar Castillo", "Icela Marissa Rodríguez Haro", "Ana Paula Vizconde Rodríguez", "Marlon Farfán Córdova", "Marco Leoncio Salazar Castillo", "Icela Marissa Rodríguez Haro", "Ana Paula Vizconde Rodríguez" ], "abstract": "Background: Plant-derived compounds are sometimes used as substitutes for pharmaceuticals. Mauritia flexuosa is a palm tree that is widely distributed in South America, especially in the Amazon region. The San Martín region of Peru, in which this species of the Arecaceae family is found, has great biological diversity and there is economic potential in the utilization of natural resources in the region. Methods: In this study, the antibacterial effect of the hydroalcoholic extract of Mauritia flexuosa leaves was evaluated for gram-positive bacteria Staphylococcus aureus ATCC 25923 and Bacillus subtilis ATCC 6633 and gram-negative Escherichia coli ATCC 25922, Pseudomonas aeruginosa ATCC 27853 and Salmonella enterica subsp. enterica ser. Typhi ATCC 11011. Mauritia flexuosa leaves were used to prepare concentrations of 10, 20, 40 and 60mg/ml. Phytochemical analysis was performed to identify secondary metabolites in the plants. For the experiment, 10 Mueller-Hinton agar plates were prepared and 1ml of bacterial inoculum, standardized to 0.5 McFarland, was added to each plate. The hydroalcoholic extract was added via the diffusion method, making five holes of 5mm each (four with extract concentrations and one with distilled water as a control group), and the plates were incubated for 24 hours at 36°C. The inhibition halo was measured in mm using a digital vernier caliper. Results: For gram-negative bacteria, an antibacterial effect was demonstrated for Pseudomonas aeruginosa only, at an extract concentration of 60mg/ml, with an inhibition halo of 14.8 mm. For gram-positive bacteria Baccillus subtilis and Staphylococcus aureus, an antibacterial effect was demonstrated at an extract concentration of 60mg/ml, with inhibition halos of 13.2mm and 15.4mm in diameter, respectively. Conclusion: It can be concluded that the hydroalcoholic extract of Mauritia flexuosa does not inhibit bacterial growth for gram-negative bacteria Salmonella Typhi and Escherichia coli.", "keywords": [ "Antibacterial", "hydro-alcoholic extract", "Mauritia flexuosa", "gram negative bacteria", "gram positive bacteria" ], "content": "Introduction\n\nFlora is important, because of its diversity, for the production of phytochemical compounds, used as a natural alternative to pharmaceuticals for the treatment of diseases caused by fungi, bacteria and other organisms1. Plant-derived compounds have been used as substitutes for pharmaceuticals; for example, one out of three people use medicinal plants for healing purposes in Europe2,3. A study conducted in the United States found that 25% of drugs are derived from plants4,5.\n\nMauritia flexuosa is a palm tree that is widely distributed in South America, especially in the Amazon region. The San Martín region of Peru, in which this species of the Arecaceae family is found, has great biological diversity and there is economic potential in the utilization of natural resources in the region. In the food industry, the shell and the endocarp of the Mauritia flexuosa fruit are usually discarded or underused for the preparation of sweets, ice creams, juices, jams, infant food and oils6. Some studies have emphasized the pharmacological potential of these parts, which contain bioactive compounds, to play a role in antimicrobial defense7–9, lipid metabolism10, hypoglycaemia11 and curative activities12. Mauritia flexuosa also produces secondary metabolites belonging to chemical groups, such as alkaloids and cyanogenic glycosides, and non-nitrogenous compounds, such as tannins, flavonoids, terpenes and anthocyanin13.\n\nBacteria have become resistant to many medicines; thus, infections have been disseminating quickly between people and animals. Gram-positive bacteria, including Staphylococcus aureus (causes a variety of infectious diseases, including skin infections, wound infections and pneumonia) and Bacillus subtilis (which is not considered a human pathogen but can cause intoxication and food contamination), as well as gram-negative bacteria, including Salmonella enterica subsp. enterica ser. Typhi (Salmonella Typhi, causes typhoid fever), Pseudomonas aeruginosa (resistant to multiple drugs and responsible for healthcare-associated infections) and Escherichia coli (causes diarrhea and renal insufficiency, which can lead to death) are involved in human bacterial infection14.\n\nMedicinal plants are considered an alternative solution for controlling bacterial diseases; thus, there is a requirement for research with the aim of finding new natural alternatives. The World Health Organization has shown that they can be effective and their percentage of health risk may be minimal15. Research shows that the hydroalcoholic extracts of leaves have an antibacterial effect on gram-positive and gram-negative bacteria because of their active principles, which play an important role in the development of new therapeutic agents16,17.\n\nThe objective of this research was to evaluate the antibacterial effect of the hydroalcoholic extract of Mauritia flexuosa leaves on gram-positive and gram-negative bacteria.\n\n\nMethods\n\nLeaves of Mauritia flexuosa were collected by researchers (ANSV, MFC, MLSC, IMRH and APVR) in the district of Cacatachi, San Martín at 295 meters above sea level, 12 km to the north of Tarapoto (6°29'40\"of South latitude and 76°27'57\" of West longitude). The specimen was transferred to the Herbarium Truxillense of Universidad Nacional de Trujillo for identification by specialists, obtaining a registration code. The sample was transported in a wooden press inside a labelled vacuum bag and kept at an ambient temperature of 37°C.\n\nWhole leaves were selected and those with signs of deterioration discarded. Leaves were washed with distilled water and disinfected with cotton dipped in 96% ethanol. Leaves were fragmented to an approximate size of 3mm and the maceration method was carried out as follows. Fresh leaves, with a weight of 200g, were washed with distilled water to remove impurities, wrapped in kraft paper and dried in a universal oven (Memmert GmbH + Co. KG) at 25 °C for approximately 12 hours. The dry sample was cut with scissors to obtain small pieces, placed in an amber glass jar and 500mL of 96% ethanol was added, followed by agitation using a vertical rotavapor (Scilogex RE-100) at 70 rpm for 10 minutes every four hours, except overnight from 10 pm to 7 am, for 15 days. The sample was filtered four times with Whatman N °1 and then N °2 filter paper. Then, the vertical rotavapor (Scilogex RE-100) was used for two hours at 70rpm to obtain a dry extract, which was dissolved in alcohol at 96 °C and used to prepare concentrations of 10, 20, 40 and 60mg/ml.\n\nThe phytochemical analysis of Mauritia flexuosa leaves was qualitative and was carried out according to the method described by Miranda and Cuellar18. Each sample was subjected to solvents of increasing polarity in order to obtain secondary metabolites according to their solubility, using reagents and dyes to determine presence or absence of active components such as terpenes, flavonoids, reducing sugars, among others. The assays used to determine the presence of each type of secondary metabolite are listed in Table 1. The results of the color change were judged by eye according to the method described by Miranda and Cuellar18 and classified as light, moderate or strong.\n\n(+), light: (++), moderate; (+++), strong.\n\nStandard bacterial strains (American Type Collection Culture (ATCC) were provided by the Bacteriology Laboratory of Universidad Nacional de Trujillo. Gram positive bacteria Staphylococcus aureus ATCC 25923 and Bacillus subtilis ATCC 6633 were used. Gram-negative Escherichia coli ATCC 25922, Pseudomonas aeruginosa ATCC 27853, and Salmonella Typhi ATCC 11011 were used.\n\nBacteria were stored at a temperature below 5°C. From there, they were removed and reactivated in brain heart infusion medium at 37°C for 18 days. Mueller-Hinton agar (Merck) was prepared in biosafety level one conditions according to the manufacturer’s specifications. 8ml of Mueller-Hinton agar was poured into 100mm Petri dishes and left to dry at 37 °C for 30 minutes. A suspension of 5×108 colony forming units (CFU) of each bacterial culture was prepared in a test tube of 10ml isotonic sodium chloride solution, equivalent to 0.5 MacFarland. For the experiment, 1ml of each suspension was spread onto Mueller-Hinton plates and was left to dry for 30 minutes. For the application of the hydroalcoholic extract, the diffusion method in agar was used19 and five holes of 5mm were made in each Petri dish; four to add 70μl of the prepared extract concentrations (10, 20, 40 and 60mg/ml) and one to add 70μl of the control (distilled water). In total, the experiment involved 50 plaques, with two Petri dishes of five plaques each used for each of the five species of bacteria. The plates were incubated at 36°C for 24 hours. The results were determined using a digital vernier caliper (CALDI-6MP, Truper), giving the diameter of the halo in mm.\n\nSPSS (version 22) software was used for the summation, averages, tables and graphs. For the calculation of the inhibition halo, the mean average of the two repetitions for each extract concentration for each bacterium was calculated.\n\n\nResults\n\nIt was observed that Mauritia flexuosa has a large number of phenolic compounds and flavonoids (Table 1)20.\n\nAs shown in Figure 1, no inhibition halos were formed in the control group (distilled water) around any of the gram-negative bacteria. For Pseudomonas aeruginosa, there was an inhibition halo of 14.8mm of diameter at a Mauritia flexuosa extract concentration of 60mg/ml, a halo of 12.4mm at 40mg/ml, a halo of 10.2mm at 20mg/ml and a halow of 8.0mm at 10mg/ml. However, for Salmonella Typhi and Escherichia coli, no inhibition halos were observed21,22.\n\nAs shown in Figure 2, the control (distilled water) did not produce inhibition halos around any of the gram-positive bacteria. For Bacillus subtilis, there was an inhibition halo with a diameter of 13.2mm at a Mauritia flexuosa extract concentration of 60mg/ml, a halo of 11.1mm at 40mg/ml, a halo of 9.4mm at 20mg/ml and a halo of 6.6mm at 10mg/ml. For Staphylococcus aureus, diameter of the inhibition halo was 15.4mm at an extract concentration of 60mg/ml, a halo of 13.2mm at 40mg/ml, a halo of 11.3mm at 20mg/ml and a halo of 8.1mm at 10mg/ml21,22.\n\n\nDiscussion\n\nMedicinal plants have become common in today's research, analyzing different parts of the plant, in order to ensure the diversity, and evaluating and their potential as therapeutic agents23–25. Therefore, the increase in bacterial genetic mutations conferring bacterial resistance to antibiotics and the increase in difficulty in treating these infections has led to the investigation of new antibacterial agents, especially those of natural origin26.\n\nAs shown in Table 1, secondary metabolites were qualitatively identified in Mauritia flexuosa leaves, such as steroids, triterpenes, phenolic compounds, flavonoids, lactones, reducing sugars and tannins. The qualitative analysis of leaves of Mauritia flexuosa is important, since the evidence that there is a large presence of secondary metabolites, such as steroids, triterpenes, phenolic compounds and flavonoids, allows comparison of its potential against growth of bacteria with other research.\n\nAs shown in Figure 1, three gram-negative bacteria were studied, of which only Pseudomonas aeruginosa was inhibited by the hydroalcoholic extract, at a concentration of 60mg/ml and producing an inhibition halo of 14.8 mm in diameter. There was no inhibitory effect on Salmonella Typhi and Escherichia coli. In a study carried out in Brazil, in which Escherichia coli ATCC 25922 and Pseudomonas aeruginosa ATCC 27853 were used, no sensitivity to the extract of Mauritia flexuosa was evidenced27. Therefore, it should be pointed out that the chemical composition of the soil and the climatic conditions may result in variation of the active constituents; therefore, differences in the results may be observed despite the use of the same bacterial species. Furthermore, it has been found previously that the hydroalcoholic extract inhibited the bacterial growth of the same species used for this study28,29. Therefore, the chemical composition of the leaves plays an essential role in the antibacterial effect, since the effect is composed of a set of active principles that work synergistically.\n\nAs shown in Figure 2, gram-positive strains of Bacillus subtilis and Staphylococcus aureus had inhibition halos of 13.2mm and 15.4mm in diameter, respectively, at a hydroalcoholic extract concentration of 60mg/ml, identifying that at a higher concentration there is lower bacterial growth. It is known that hydroalcoholic extracts of plants release a large quantity of phenols and flavonoids and other compounds, including great diversity of secondary metabolites, which may explain this antibacterial action30–32. Similar results were identified in other studies that tested hydroalcoholic extracts on gram-positive bacteria, finding that the extract had an antibacterial effect28,33,34. Therefore, it is important to emphasize the importance of each active principle since act synergistically to obtain better results, being more efficient and effective35. Finally, gram-positive and gram-negative bacteria can be pathogenic; hence, the importance of identifying new alternatives to mitigate the pathologies caused by these bacteria36. Therefore, it is necessary to study regional plants as a treatment alternative and develop new plant products with protective effects.\n\nIn conclusion, the hydroalcoholic extract of Mauritia flexuosa does not inhibit bacterial growth of gram-negative bacteria Salmonella Typhi and Escherichia coli. However, in Pseudomonas aeruginosa the extract inhibits growth at a concentration of 60mg/ml, forming an inhibition halo of 14.8 mm. For gram-positive bacteria Bacillus subtilis and Staphylococcus aureus, bacterial inhibition is observed at a concentration of 60mg/ml, with halos of 13.2mm and 15.4mm of diameter, respectively, demonstrating an antibacterial effect in these species.\n\n\nData availability\n\nFigshare: Supporting data for Figure 1 and Figure 2. https://doi.org/10.6084/m9.figshare.8051870.v321\n\nFigshare: Photographs of each agar plate. https://doi.org/10.6084/m9.figshare.8312777.v322\n\nPhotographs of the phytochemical analyses. https://doi.org/10.6084/m9.figshare.8862431.v120\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nChilquillo Torres EA, Albán J, Muñoz y A: Estudio etnobotánico de plantas medicinales utilizadas en comunidades adyacentes al Área de Conservación Privada San Antonio, Chachapoyas, Amazonas, Perú. Revista de Investigación Científica UNTRM: Ciencias Naturales e Ingeniería. 2018; 1(1): 65–73. Reference Source\n\nGallegos M: Las plantas medicinales: principal alternativa para el cuidado de la salud, en la población rural de Babahoyo, Ecuador. Anales de la Facultad de Medicina. 2016; 77(4): 327–32. Publisher Full Text\n\nRodríguez N, Pérez J, Iglesias J, et al.: Actualidad de las plantas medicinales en terapéutica. Acta Farmacéutica Portuguesa. 2015; 4(1): 42–52. Reference Source\n\nCorrales I, Reyes C: Actividad etnofarmacológica y antimicrobiana de los componentes químicos de las plantas medicinales utilizadas en Estomatología. 2015; 54(257): 71–83. Reference Source\n\nHeisler E, Budó M, Schimith M, et al.: Uso de plantas medicinais no cuidado á saúde: producáo científica das teses e dissertacóes da enfermagem brasileira. Enfermería Global. 2015; 14(3): 390–417. Reference Source\n\nLins Chaves T, Ricardo L, de Paula-Souza J, et al.: Useful Brazilian plants under the view of the writer-naturalist João Guimarães Rosa. Rev Bras Farmacogn. 2015; 25(5): 437–444. Publisher Full Text\n\nKoolen H, Da Silva F, Gozzo F, et al.: Antioxidant, antimicrobial activities and characterization of phenolic compounds from buriti (Mauritia flexuosa L. f.) by UPLC–ESI-MS/MS. Food Res Int. 2013; 51(2): 467–473. Publisher Full Text\n\nKoolen HH, Soares ER, da Silva FM, et al.: Mauritic acid: a new dammarane triterpene from the roots of Mauritia flexuosa L.f. (Arecaceae). Nat Prod Res. 2013; 27(22): 2118–2125. PubMed Abstract \| Publisher Full Text\n\nSiqueira EP, Andrade A, Souza E, et al.: In vitro antibacterial action on methicillin-susceptible Staphylococcus aureus (MSSA) and methicillin-resistant (MRSA) and antitumor potential of Mauritia flexuosa L. f. Journal Medicinal Plants Research. 2014; 8(48): 1408–1417. Reference Source\n\nAquino Jde S, Tavares RL, Medeiros Lde B, et al.: Effect of the consumption on buriti oil on the metabolism of rats induced by iron overload. Arch Endocrinol Metab. 2015; 59(5): 422–7. PubMed Abstract \| Publisher Full Text\n\nBataglion G, Da Silva F, Eberlin M, et al.: Simultaneous quantification of phenolic compounds in buriti fruit (Mauritia flexuosa). Food Res Int. 2014; 66: 396–400. Publisher Full Text\n\nBatista J, Olinda R, Medeiros V, et al.: Atividade antibacteriana e cicatrizante do óleo de buriti Mauritia flexuosa L. Ciencia Rural. 2012; 42(1): 136–141. Publisher Full Text\n\nPereira Freire JA, Barros KBNT, Lima LKF, et al.: Phytochemistry Profile, Nutritional Properties and Pharmacological Activities of Mauritia flexuosa. J Food Sci. 2016; 81(11): R2611–R2622. PubMed Abstract \| Publisher Full Text\n\nSika C, Sina H, Adoukonou H, et al.: Antimicrobial activity of Anacardium occidentale L. leaves and barks extracts on pathogenic bacteria. Afr J Microbiol Res. 2014; 8(25): 2458–2467. Publisher Full Text\n\nOrganización Mundial de la Salud: Estrategia de la OMS sobre medicina tradicional 2014-2023. Reference Source\n\nJan S, Khan MR: Protective effects of Monotheca buxifolia fruit on renal toxicity induced by CCl4 in rats. BMC Complement Altern Med. 2016; 16(1): 289. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGouveia-Figueira SC, Gouveia C, Carvalho MJ, et al.: Antioxidant Capacity, Cytotoxicity and Antimycobacterial Activity of Madeira Archipelago Endemic Helichrysum Dietary and Medicinal Plants. Antioxidants (Basel). 2014; 3(4): 713–29. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMiranda M, Cuellar A: Mención en Productos Naturales y Terapéuticos. Maestría en Farmacia y Bioquímica. Escuela de Post-Grado. Universidad Nacional de Trujillo. Cuba, 2003; 3–5.\n\nStella L, Marín D: Metodologías para evaluar in vitro la actividad antibacteriana de Compuestos de or ígen vegetal. Scientia et technical. Universidad Tecnol ógica de Pereira. 2009; 42(1). Reference Source\n\nSandoval A: Photochemical analyses for Mauritia flexuosa. 2019. http://www.doi.org/10.6084/m9.figshare.8862431.v1\n\nSandoval A: Supporting data for Figure 1 and Figure 2. 2019. http://www.doi.org/10.6084/m9.figshare.8051870.v3\n\nSandoval A: Photos of agar plates (unedited). 2019. http://www.doi.org/10.6084/m9.figshare.8312777.v3\n\nAfsar T, Khan MR, Razak S, et al.: Antipyretic, anti-inflammatory and analgesic activity of Acacia hydaspica R. Parker and its phytochemical analysis. BMC Complement Altern Med. 2015; 15: 136. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBokhari J, Khan MR: Evaluation of anti-asthmatic and antioxidant potential of Boerhavia procumbens in toluene diisocyanate (TDI) treated rats. J Ethnopharmacol. 2015; 172: 377–85. PubMed Abstract \| Publisher Full Text\n\nUllah S, Rashid Kha M, Ali Shah N, et al.: Ethnomedicinal plant use value in the Lakki Marwat District of Pakistan. J Ethnopharmacol. 2014; 158 Pt A: 412–22. PubMed Abstract \| Publisher Full Text\n\nMiyasaki Y, Rabenstein JD, Rhea J, et al.: Isolation and characterization of antimicrobial compounds in plant extracts against multidrug-resistant Acinetobacter baumannii. PLoS One. 2013; 8(4): e61594. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDe Oliveira AI, Mahmoud TS, do Nascimento GN, et al.: Chemical Composition and Antimicrobial Potential of Palm Leaf Extracts from Babaçu (Attalea speciosa), Buriti (Mauritia flexuosa), and Macaúba (Acrocomia aculeata). ScientificWorldJournal. 2016; 2016: 9734181. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBandara KRV, Padumadasa C, Peiris DC, et al.: Potent antibacterial, antioxidant and toxic activities of extracts from Passiflora suberosa L. leaves. PeerJ. 2018; 6: e4804. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEspadero M, Armijos L, Ávila L, et al.: Microbiological Evaluation And Chemical Composition Of Organic Extracts From Euphorbia aff. viridis (Klotzsch & Garcke) Boiss ON Staphylococcus Aureus, Klebsiella Pneumoniae AND Escherichia Coli. Revista La Granja de Ciencias de la Vida. 2019; 29(1): 119–129. Publisher Full Text\n\nRavi B, Ramachandra Y, Sundara S, et al.: Evaluating the anthelmintic potential of leaf gall extracts of Terminalia chebula (Gaertn.) Retz. (Combretaceae). Journal of Young Pharmacists. 2014; 6(3): 1–4. Reference Source\n\nEshwarappa RS, Iyer S, Subaramaihha SR, et al.: Antioxidant activities of ficus glomerata (moraceae) leaf gall extracts. Pharmacognosy Res. 2015; 7(1): 114–20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEshwarappa R, Ramachandra YL, Subaramaihha SR, et al.: Antioxidant activities of leaf galls extracts of Terminalia chebula (Gaertn.) Retz. (Combretaceae). Acta Sci Pol Technol Aliment. 2015; 14(2): 97–105. PubMed Abstract \| Publisher Full Text\n\nde Oliveira AM, Mesquita Mda S, da Silva GC, et al.: Evaluation of Toxicity and Antimicrobial Activity of an Ethanolic Extract from Leaves of Morus alba L. (Moraceae). Evid Based Complement Alternat Med. 2015; 2015: 513978. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAlvarado S, Plasencia P, Miñano E, et al.: Efecto antibacteriano in vitro del extracto etanólico de Prosopis pallida sobre Enterococcus faecalis ATCC 29212. Revista Cubana de Medicina Tropical. 2018; 70(2). Reference Source\n\nHenciya S, Seturaman P, James AR, et al.: Biopharmaceutical potentials of Prosopis spp. (Mimosaceae, Leguminosa). J Food Drug Anal. 2017; 25(1): 187–96. PubMed Abstract \| Publisher Full Text\n\nSika C, Sina H, Adoukonou H, et al.: Antimicrobial activity of Anacardium occidentale L. leaves and barks extracts on pathogenic bacteria. Afr J Microbiol Res. 2014; 8(25): 2458–2467. Publisher Full Text" }	[ { "id": "52932", "date": "27 Sep 2019", "name": "Flor de María Evangelista Montoya", "expertise": [ "Reviewer Expertise Microbiology", "antibiotic resistance." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nSandoval and collaborators investigated the antibacterial effect of the hydroalcoholic extract of Mauritia flexuosa leaves on gram-negative and gram-positive bacteria. They tested this effect on standard strains of 3 gram-negative and 2 gram-positive bacteria using the agar diffusion method. They found that the extract had antibacterial effects against the pathogens S. aureus, B. subtilis and P. aeruginosa but not against E. coli and S. typhi. Inhibition was observed at 60 mg/mL (4.2 mg/well).\nOverall the article is clearly written, however it could benefit from some English grammar and style editing.\nAlthough the Introduction expresses the present art well, I suggest including a short explanation of why the authors decided to use the leaves of M. flexuosa for their study. Additionally, references 16 and 17 do not seem appropriate to support the use of leaves because they deal with fruit and total aerial parts.\nIn the Methods section authors say: \"The sample was transported in a wooden press inside a labelled vacuum bag and kept at an ambient temperature of 37˚C \". Usually ambient temperature is approximately 25˚C. I request the authors clarify the temperature used.\nWhy was it necessary to reactivate the bacterial strains for 18 days? That seems like a very long time, please clarify.\nThe reference used (18) to describe the method used for the phytochemical analysis of the extract is not easily accessible to the reader. I recommend referencing a standard review or book in which these common procedures are explained.\n\nIn the Results section, the authors show that the extract has antibacterial effect against some bacteria. This is useful data as it may point to new leads for antibiotics. However, it seems that there are some deficiencies in proper experimental controls. First, distilled water was used as a negative control, but the extract was dissolved in ethanol. Therefore, it is possible that some residual ethanol may be causing a bacteriostatic effect. Second, no positive control with a standard antibiotic was performed. This makes it difficult to make clear comparisons of the antibiotic activity of the extract to known antibiotics and to establish what halo size determines sensitivity.\n\nThe discussion expresses the importance of this type of study and makes some comparisons of the present results with other studies. However, the way it is written can mislead the reader to the idea that any hydroalcoholic extract from any plant has an antibacterial effect against the bacteria tested and does not highlight the importance of using the extract from M. flexuosa leaves for antimicrobial use. I recommend including a statement and explanation about this.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4978", "date": "21 Oct 2019", "name": "ANA SANDOVAL VERGARA", "role": "Author Response", "response": "Dear.Flor de María Evangelista Montoya.I appreciate your comments, below I detail each suggestion:1. M. flexuosa it is one of the most consumed fruit species in the Peruvian Amazon, the inhabitants use by-products such as leaves, fruits, seeds to cure diseases; Therefore, the importance of its study as an alternative to traditional medicine.The references indicated were considered by the authors since it is a plant extract, they refer to the importance of the active ingredients and the antibacterial effect they possess; Starting from that premise is a valuable criterion to demonstrate the presence of toxic components in plants that could be an option for the treatment of diseases caused by bacteria.2. The temperature at which the plant samples were transported was 25 ° C.3. The bacteria mentioned only reactivate in 18 hours.4. Another author who clearly describes the phytochemical march in plants has been considered.Lock, O. Phytochemical Research: Methods in the study of natural products. 3rd edition, Editorial of the Department of Sciences. Pontifical Catholic University of Perú, 2016: p 6-8 5. First. - When the ethanol extract is prepared, the ethanol is completely removed by evaporation, therefore it is not possible to find residual alcohol consequently does not cause bacteriostatic effectSecond. - The antibiotic is a drug synthesized and industrially purified, while the plant extract has a set of naturally produced drugs and the authors did not consider it as a positive control, since comparisons cannot be made only serve as a reference.6. This study contributes to the knowledge of the species and the expansion of its application in biotechnology, as well as the results obtained contribute to a better understanding of the relationship between the chemical composition of the hydroalcoholic leaf extract and its antibacterial potential, this plant is widely used in the food industry, taking into account the diversity of its active ingredients, future research should be encouraged and the main objective should be to isolate specific compounds to produce safe phytotherapies" } ] }, { "id": "59415", "date": "02 Mar 2020", "name": "Mario R. Esparza Mantilla", "expertise": [ "Reviewer Expertise Microbiology", "Biotechnology", "Environmental Microbiology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe research proposes to demonstrate the antimicrobial effect of a natural extraction of plants, this is well known and the test models also, in that direction to work it requires more evidence or generate new contributions to change the models or use more robust ones for the tests, perhaps the systems of broths of cultures in 96-plate plates and evaluate the effect in vivo allow a result of greater impact, on the other hand the controls applied to the experiment such as water are not so blunt should the solvent with which the plant extract.\nOn the application of more robust methods, I suggest applying the MIC-MBC (minimum inhibitory concentration (MIC) and minimum bactericidal concentration (MBC), the methodology recommended by the Clinical and Laboratory Standards Institute), as well as the technique of viable microdrops. It is also suggested to construct microbial kinetic curves to have a more precise data of the effect of the dose vs the bacterial cell density in a certain time and can be contributed in the knowledge of microbial physiology since under these conditions the extract could also have another form of action closer to the real one (within the host) that provides an approach to results that allow a future biomedical application of the plant extract.\nResearchers should improve the conclusion since their writing only reports one result, they should be able to generate new knowledge based on the analysis and discussion of the results that will allow a substantive contribution to the execution of the proposed research.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-1487
https://f1000research.com/articles/7-1828/v1	21 Nov 18	{ "type": "Case Report", "title": "Case Report: Double penile fracture", "authors": [ "Felipe Mercado-Olivares", "J. Antonio Grandez-Urbina", "Giomar Farfan-Daza", "Juan Pacheco-Sauñe", "Luciano Nuñez-Bragayrac", "Felipe Mercado-Olivares", "Giomar Farfan-Daza", "Juan Pacheco-Sauñe", "Luciano Nuñez-Bragayrac" ], "abstract": "Penile fracture is an underreported surgical emergency. It usually occurs as a single rupture of the tunica albuginea in one of two corpora cavernosa; a rupture of both masses is an uncommon finding. We report a case of a young male who presented to the emergency department two hours after sustaining penile trauma. Prompt surgical exploration was performed four hours post-injury. He was found to have one fracture on each corpora cavernosa, without urethral injury, which were repaired successfully. The patient had a favorable recovery and was discharged on the third postoperative day without complications. The aim of this report is to highlight the importance of complete degloving of the penile shaft for a meticulous search during surgery to avoid missed injuries. This approach will ensure a successful outcome avoiding physical and psychological disabilities.", "keywords": [ "Penile Diseases", "Urologic Surgical Procedures", "Injuries", "Penile induration" ], "content": "Introduction\n\nFracture of the penis, or faux pas du coit, is an uncommon surgical emergency that occurs as a result of trauma to the erect penis1,2. The incidence of penile injuries is underreported because many patients do not seek medical attention3. Penile fracture is defined as the rupture of the tunica albuginea and the corpora cavernosa4. Most cases occur during sexual intercourse, usually due to hitting the symphysis pubis or the perineum after the penis slips out of the vagina; less commonly reported is during masturbation2,3.\n\nIt is manifested by a cracking or popping sound accompanied by immediate detumescence, followed by rapid swelling, widespread ecchymosis, sharp pain and deformity (away from the trauma site)2,5.\n\nThis condition can be quickly diagnosed after history taking, physical examination and imaging. Prompt diagnosis and early surgical repair are essential to ensure a successful outcome6,7.\n\nIt is uncommon for a penile fracture to involve both of the corpora cavernosa. We report a rare case of a double penile fracture and describe its presentation and management.\n\n\nCase report\n\nA 32-year-old patient presented to the emergency department two hours after having penile trauma during sexual intercourse. He was having sexual intercourse in a 'woman-on-top' position when he heard a ‘snap’ sound followed by severe pain and immediate loss of tumescence. On physical examination, his penis was swollen, deformed, and with signs of ecchymosis. On admission, two ventral irregularities were found during penile palpation (Figure 1). There was no urethral bleeding, nor voiding difficulties suggesting an uncompromised urethra. During the patient's consultation, no previous sexual disorders were found and the patient denied the usage of PDE5 inhibitors or intra-carvernosum injections. Labs results were within normal limits and radiological images were not obtained prior to surgical intervention due to lack of resources.\n\nCefazoline 2gr IV was administered prophylactically before surgery. Emergency exploration was performed 4 hours post-penile injury. A 16 Fr. Foley catheter was placed without difficulties. Then a circumcision was made, and surgical exploration was performed. On degloving the penile skin, a fracture on each corpora cavernosa was found. The length of the right and left defects were 25mm and 35mm, respectively (Figure 2). In order to protect the urethra, a Penrose drain was used as a vessel loop through the double fracture separating it from the site of injury (Figure 3). Repair of the two lacerations was done using 3-0 Vicryl in a continuous suture pattern. The Foley catheter was removed the next day without adverse events. The patient had a favorable recovery and was discharged three days after the surgical procedure without any complication (Figure 4).\n\nAt a follow-up period of 90 days, the patient had a proper erectile function, no deformity, no pain during sexual intercourse, and no voiding difficulties.\n\n\nDiscussion\n\nThe penis is composed of erectile tissue arranged in a columnar fashion. Two dorsolateral corpora cavernosa and one ventral corpus spongiosum, each enclosed by the tunica albuginea. The urethra traverse the corpus spongiosum throughout its length. The distal expansion of the corpus spongiosum forms the glans penis. Buck’s fascia encloses the corpus spongiosum ventrally, and splits dorsally to surround the two corpora cavernosa1,4,5.\n\nPenile fracture, or faux pas du coit, is an underreported but emergent urological condition8. It’s underreported because many patients do not seek medical attention due to embarrassment, shame, or lack of guidance3. Activities that can result in penile fracture includes self-manipulation to achieve tumescence, sexual intercourse, turning over in bed and a direct blow to the erect penis2,4. The most common cause is sexual intercourse, with injuries often caused by different sexual positions. In an original article published in 2017, Barros et al. reported that ‘doggy style' was more commonly associated with double fractures9. In fact, any activity associated with tumescence can increase the chance of penile fractures. In 2002, Blake et al. reported the first case of a fracture related to pharmacologically induced erections10. The increased risk of penile fractures during tumescence is related to the tunica albuginea stretching and thinning. The thickness of the tunica reduces from 2.4mm, in the flaccid state, to 0.25-0.5mm in the erect form 11.\n\nThe most extensive review of cases was made by Eke between 1935 and 2001. He analyzed 1331 cases from 183 publications. Most reports were from the Mediterranean region; the median age of patients were 35 years. Clinical features included sudden penile pain, detumescence, voiding difficulties, penile swelling and deviation. Associated injuries included urethral rupture. Complications of the rupture included coital difficulty, urethral fistula, penile plaque and erectile dysfunction1.\n\nThe diagnosis is mainly clinical, although in the absence of typical signs it can pose a real challenge. Therefore, an adequate history and physical examination are cornerstones of the diagnostic process. In our patient, radiologic investigations were not performed due to lack of resources and the diagnosis was based solely on physical examination findings. Most cases reported in literature describes imaging modalities being used to exclude the presence of a concomitant urethral injury and to delineate the exact location of the albuginea rupture12. Various imaging modalities have been used to aid in the diagnosis, such as cavernosography, retrograde urethrography, ultrasonography-colour Doppler, and magnetic resonance. Cavernosography is an invasive method that is rarely used. Retrograde urethrography is also an invasive method, but is required only if associated urethral injury is suspected7. Ultrasonography is the modality most frequently used because of its cost and availability. Magnetic resonance is the most accurate method to localize the lesions, but its availability is limited due to its cost13.\n\nMost studies report a solitary fracture of the corpora cavernosa. In our patient, both corpus were found to have fractures. Although no imaging modalities were performed, the omission of careful surgical exploration for the second site of injury might have led to unsatisfactory outcomes. Therefore, it seems essential to make a complete surgical exploration, degloving the penile shaft, while operating upon such cases.\n\nUrgent surgery is the recognized gold standard approach. However, recent studies have revealed that there is no statistical difference in long-term outcomes of early versus delayed repair in patients without urethral involvement14,15. In 2011, Kozacioglu et al. evaluated 56 patients who underwent early and delayed penile fracture repair. Their study noted no serious deformities nor erectile dysfunction in the long term as a result of a delay in surgery in patients without urethral involvement5. In our case, surgical repair was offered early in the course of the presentation. The outcome was favorable and no complications were noted on follow-up.\n\n\nConclusion\n\nDouble rupture of the tunic albuginea is a rare finding in penile fractures. The diagnosis is clinical; however imaging should be used to help delineate the exact location and extent of injury. Degloving of the entire penis is recommended to identify possible multiple fractures, urethral damage, and damage to nearby structures. In our case, surgical repair was performed early which led to a successful recovery without complications. This case report is intended to further increase the literature helping physicians in search of experiences managing this low incidence condition.\n\n\nConsent\n\nWritten informed consent for publication of their clinical details and/or clinical images was obtained from the patient.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.", "appendix": "Grant information\n\nThe author(s) declared that no grant(s) were involved in supporting this work.\n\n\nReferences\n\nEke N: Fracture of the penis. Br J Surg. 2002; 89(5): 555–65. PubMed Abstract\n\nFergany AF, Angermeier KW, Montague DK: Review of Cleveland Clinic experience with penile fracture. Urology. 1999; 54(2): 352–5. PubMed Abstract \| Publisher Full Text\n\nKrishna Reddy SV, Shaik AB, Sreenivas K: Penile injuries: A 10-year experience. Can Urol Assoc J. 2014; 8(9–10): E626–31. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKachewar S, Kulkarni D: Ultrasound evaluation of penile fractures. Biomed Imaging Interv J. 2011; 7(4) e27. [citado 9 de octubre de 2018]. PubMed Abstract \| Free Full Text\n\nKozacioglu Z, Degirmenci T, Arslan M, et al.: Long-term significance of the number of hours until surgical repair of penile fractures. Urol Int. 2011; 87(1): 75–9. PubMed Abstract \| Publisher Full Text\n\nNasser TA, Mostafa T: Delayed surgical repair of penile fracture under local anesthesia. J Sex Med. 2008; 5(10): 2464–9. PubMed Abstract \| Publisher Full Text\n\nDhillan R, Sen D: Emergency surgical management of double penile fracture at a peripheral hospital. Med J Armed Forces India. 2015; 71(Suppl 1): S46–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBoncher NA, Vricella GJ, Jankowski JT, et al.: Penile fracture with associated urethral rupture. Case Rep Med. 2010; 2010: 791948. [citado 9 de octubre de 2018]. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBarros R, Schulze L, Ornellas AA, et al.: Relationship between sexual position and severity of penile fracture. Int J Impot Res. 2017; 29(5): 207–9. PubMed Abstract \| Publisher Full Text\n\nBlake SM, Bowley DMG, Dickinson A: Fractured penis: another complication of sildenafil. 2. Reference Source\n\nBitsch M, Kromann-Andersen B, Schou J, et al.: The elasticity and the tensile strength of tunica albuginea of the corpora cavernosa. J Urol. 1990; 143(3): 642–5. PubMed Abstract \| Publisher Full Text\n\nFalcone M, Garaffa G, Castiglione F, et al.: Current Management of Penile Fracture: An Up-to-Date Systematic Review. Sex Med Rev. 2018; 6(2): 253–60. PubMed Abstract \| Publisher Full Text\n\nChung CH, Szeto YK, Lai KK: 'Fracture' of the penis: a case series. Hong Kong Med J. 2006; 12(3): 197–200. PubMed Abstract\n\nNaraynsingh V, Hariharan S, Goetz L, et al.: Late delayed repair of fractured penis. J Androl. 2010; 31(2): 231–3. PubMed Abstract \| Publisher Full Text\n\nWong NC, Dason S, Bansal RK, et al.: Can it wait? A systematic review of immediate vs. delayed surgical repair of penile fractures. Can Urol Assoc J. 2017; 11(1–2): 53–60. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "46660", "date": "25 Apr 2019", "name": "Rodrigo Barros", "expertise": [ "Reviewer Expertise Sexual medicine" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article describes a case of complex penile fracture (PF) with bilateral corpus cavernosum involvement resulting from sexual activity. Sexual intercourse represents the main cause of PF and is usually associated with high-energy traumas causing double fractures. This may occur during more vigorous sexual intercourse, especially in certain positions. According to our previous study, “doggy style” position was most frequently associated with complex PF, second only to the “man-on-top” position. Differently from this case report, in our sample, the “woman-on-top” position showed a low incidence, and no significant association with the severity of the PF. The authors should mention some aspects that can lead to complex bilateral lesions, such as vigorous sexual intercourse and certain types of sexual positions.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Partly\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [] }, { "id": "48955", "date": "04 Jun 2019", "name": "Sameer Trivedi", "expertise": [ "Reviewer Expertise Andrology", "Uro-oncology", "Endourology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article has few grammatical errors and needs better proof reading. Authors fail to mention other aetiological factors like the practice of “Taqaandan” in Middle Eastern countries. Authors have stated that involvement of both cavernosa is uncommon in penile fractures but have failed to mention the incidence in absolute numbers. The authors did not perform any radiological imaging study because of lack of resources. It implies that they would have preferred to perform some imaging procedure had the required resources been available. It would be interesting to know what imaging modality they would have preferred to use in such a clinical scenario. The authors state that absence of urethral bleeding and voiding difficulties was an indicator of uncompromised urethra. However, it would be prudent to reiterate that both of these factors are not reliable in ruling out associated urethral injuries as an overlying clot can mask the urethral disruption. Even a retrograde urethrogram can be falsely normal in such settings. A high degree of suspicion, complete exposure of corpus spongiosum and use of intra-operative manoeuvres like urethral injection of methylene blue have been advocated to detect occult urethral injuries. In discussion, authors have mentioned that there is no statistical difference in long-term outcomes of early versus delayed repair in patients without urethral involvement (Ref 14, 15). However, the quoted study by Naraynsingh et al (ref 14) is a case report, without any statistical analysis. In conclusion, authors have stated that imaging should be used to delineate the location and extent of the injury. However, as per both AUA and EAU guidelines, role of imaging is largely limited to unclear cases with equivocal findings or unreliable history (Ref below in citations).\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Partly\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Partly\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1828
https://f1000research.com/articles/8-290/v1	14 Mar 19	{ "type": "Research Note", "title": "Assessing knowledge of genetics in undergraduate students in Quito, Ecuador", "authors": [ "David Ortega-Paredes", "César Larrea-Álvarez", "Michelle Herrera", "Esteban Fernandez-Moreira", "Marco Larrea-Álvarez", "César Larrea-Álvarez", "Michelle Herrera" ], "abstract": "Knowledge of genetics is crucial for understanding genetic and genomic tests and for interpreting personal genomic information. Despite this relevance, no data are available about the level of knowledge of genetics in an Ecuadorian population. This investigation sought to survey such knowledge in undergraduate students affiliated with private and public institutions in Quito, the capital city of Ecuador. A total of 350 individuals responded to a validated questionnaire measuring knowledge of genetics. Scores ranged from 45% to 87% (mean: 66.8%), and students achieved slightly better results when asked about genetics and diseases (mean score: 68.3%) than when asked about genetic facts (mean score: 64.9%). Additionally, no significant differences in performance were found among students from private and public institutions. Surprisingly, the lower score obtained (45%) was from a question about how chromosomes are passed to the next generation. The highly educated status of the surveyed population could explain the overall adequate results; nonetheless, the possibility that the correct responses were given by chance cannot be ignored. Therefore, the actual knowledge of genetics among the participants might be less than that revealed by the percentages of correct answers. Consequently, to achieve the goal of ensuring informed decision-making concerning genetic and genomic tests, it seems evident that the national education programs of Ecuador require improvement in teaching of genetic concepts.", "keywords": [ "Ecuador", "knowledge of genetics", "genetic literacy", "undergraduate students", "survey" ], "content": "Introduction\n\nGenetic and genomic testing have transformed our understanding of our health, personal well-being and recreational consumerism. Advances in powerful and cheap genetic analyses have allowed new opportunities to generate information about important conditions, such as cancer, diabetes, and cardiovascular diseases (Burton, 2015; Perkins et al., 2018; Rafiq et al., 2015; Roberts & Middleton, 2018; Wu et al., 2019). In recent years, access to pharmacogenomics, nutrigenomics, disease risk, ancestry and ethnicity tests, as well as access to sport genetic analyses, has become widespread in low- and middle-income countries. Such genetic and genomic practices are carried out by health care institutions and, moreover, direct-to-consumer (DTC) genetic tests are easily available on the internet (Covolo et al., 2015; Phillips, 2016). In Ecuador, a case study by the Red Cross found that rape, intimate partner violence and femicide rates are high. Ecuadorian laws offer mothers the right to ask for a free paternity test; a positive result automatically obliges fathers to provide support for their children. Additionally, genetic tests are routine in Ecuador for police investigating rape cases. For these reasons, increasing knowledge about how DNA can be a link between parents and children or between a sexual offender and a crime seems to be a powerful tool for women’s empowerment. Several studies have demonstrated that the understanding and interpretation of personal genomic information is biased by one’s own knowledge and appreciation of basic genetic facts, namely, their level of genetic literacy (Hooker et al., 2014; Lea et al., 2011; Lontok et al., 2015; Rafiq et al., 2015). Evidently, an adequate amount of basic knowledge about genetics is essential to understand and interpret the results of genetic and medical analyses. Therefore, various studies have focused on assessing the impact of knowledge of genetics on perception of genetic facts and understanding of disease onset (Haga et al., 2013; Hollands et al., 2016; Lea et al., 2011). Unfortunately, despite the obvious necessity to determine knowledge of genetics, to our knowledge there is no available information regarding this matter in our country. Moreover, recent research has shown differences in quality between public and private higher education institutions in Colombia (Cayon et al., 2017). Therefore, it seems important to assess such differences in Ecuador. The data gathered from these kinds of studies could contribute to the development of programs to reinforce the teaching of genetics to a wider population, which will undoubtedly have a positive impact on national educational programs. Therefore, as a baseline report, we decided to determine the basic knowledge of genetics in undergraduate students in Quito, the capital city of Ecuador. This study provides a glimpse of student perspectives toward genetics and the relation of genetics to disease in a relatively highly educated population based in a developing country. Furthermore, this investigation represents one of the first steps required for building the appropriate strategies to comprehensively assess knowledge of genetics and to ultimately increase the level of genetic literacy in the region.\n\n\nMethods\n\nParticipants’ knowledge of genetics was measured as a baseline report on the attitude among undergraduate students toward genetic concepts, who were intentionally chosen because they did not follow programs involving biologically related courses (n=350 by convenience sampling method). The main objective of this research was to assess the competence of students to respond to a validated survey evaluating a minimum, adequate amount of knowledge about genetics (Fitzgerald-Butt et al., 2016). Surveys were carried out from August to October 2018. Individuals were recruited from 3 public and 4 private institutions located in Quito, the capital city of Ecuador. The identity of the institutions was handled in an anonymous form. The participants were approached at random inside the campuses and asked to fill out a questionnaire consisting of 18 statements, provided in Dataset 1 (Larrea, 2019), which measured both the actual knowledge of the associations of genetic conditions with diseases and the actual knowledge of genetic facts. For each question, the results are presented as the percentage of correct answers.\n\nPearson’s chi-square test was used to determine the likelihood that the results (answers) supporting the null hypothesis are not due to chance. Additionally, Student’s t-test was used to assess whether the two groups, composed of publicly and privately educated students, presented any significant differences regarding their measure of knowledge about genetics (assuming equal variances). P values are reported using a Type I error level of 0.05, 0.01 and 0.001. All data analyses were carried out with MATLAB® version 9.9.9341360 (R2016a). A MATLAB script to repeat the analysis is available in Dataset 2 (Larrea, 2019).\n\nThis survey was performed under the format of “common social topics”. Because of the low-risk nature of the study, approval from a committee was not sought. The participants were informed about the objective of the questionnaire; the survey was voluntary and anonymous, and information that could put the person at risk was not collected. All surveyed students provided prior verbal consent. Written consent was not sought from the participants due to the low-risk nature of the study.\n\n\nResults\n\nIn this research, we present the data gathered as a reference study outlining the knowledge of genetics in undergraduate students. Overall, 350 participants were enrolled in this research (average age: 21.8 years old, SD: ± 2.8); individuals came from diverse backgrounds that did not involve life sciences or medicine. The results varied from 45% to 87% (mean: 66.8%, median: 65%) (Table 1). The responses to each question can be found in Dataset 3 (Larrea, 2019). The percentage scores were higher for the subsection regarding the relationship between genetics and the presence of illness (mean: 68.3%). The lower scores within this section were observed when individuals were asked about the inheritance of diseases (mean: 56%, p=0.019) and when questioned about the health status of a person carrying an altered gene (mean: 55%, p=0.069). The percentage scores were lower for the subsection regarding genetic facts (mean: 64.9%). In particular, the students seemed to have difficulty answering correctly when asked about the quantity of chromosomes present in humans (mean: 58%, p=0.004) and about the number of copies of each chromosome passed down to the next generation (mean: 45%, p=0.054). In addition to the lower scores, the hypothesis that the questions were answered correctly without any previous knowledge (provided by chance) could not be significantly rejected. Generally, no differences in the overall knowledge of genetics could be found among students enrolled in private and public institutions (p=0.9405). Likewise, no differences between these two groups were observed regarding disease-related questions (p=0.7844) and genetic facts (p=0.7318).\n\nap-values for determining answers provided by chance were calculated using Pearson’s Chi squared test. T, true; F, false.\n\n\nDiscussion\n\nIn this report, we portray the percent of correct answers to an 18-item questionnaire measuring a minimum, adequate amount of knowledge about genetics. Overall, this Andes-located population of undergraduate students demonstrated some basic knowledge toward genetic concepts and their relation to diseases. Nonetheless, student knowledge on facts about genetic proved to be less strong. This tendency was observed in both privately and publicly educated individuals with no significant difference. These results are lower in comparison to the published reports on general populations that have made extensive use of similar survey instruments to determine knowledge about genetics. For instance, Haga & colleagues (2013) found higher scores in a general population based in the US. However, similar scores to those reported here were found by Jallinoja & Aro (1999) in a study performed on a general population in Finland. Furthermore, a group composed of adolescents and young adults suffering from congenital heart disease scored similar results (Fitzgerald-Butt et al., 2016). Notably, the present results are somewhat higher than those obtained from a Dutch population suffering from asthma, diabetes mellitus type II and cardiovascular disease (Calsbeek et al., 2007). It is evident that demographic differences may account for the variances in the results. Nevertheless, these results may also imply notable differences between Ecuadorian, US and European science and health education programs (Lontok et al., 2015). To our surprise, the lowest scores obtained were for the two questions involved in how chromosomes are passed down to the next generation. This means that students may not understand the power of genetics to address important issues for the Ecuadorian population, such as determining paternity, solving crimes or understanding our ethnic genetic background. To the best of our knowledge, this study is the first to report a measurement of knowledge of genetics in an Ecuadorian population.\n\nAdditionally, the presented results indicate that, despite the relatively adequate scores, the probability that participants were providing correct responses by chance could not be significantly discarded for the overall test and for the two subsections (Table 1). This fact implies that the actual knowledge might be truly weaker than the one asserted by the percentages of correct answers. However, individuals affiliated with private and public universities responded with similar accuracy. The observed average scores might reflect the high level of education of the respondents. It is worth mentioning that the interviewed people did not follow any biologically/medically related courses, which indicates that a non-specialized population may have adequate knowledge about the essential genetic concepts and their involvement in disease. Nevertheless, the participants’ knowledge may not be as strong as it appears. As mentioned earlier, the scores do not differ substantially from the earlier studies making use of similar surveys. Nonetheless, the scores were lower than those obtained from a study performed in the US (Haga et al., 2013) where genetic education is constantly improving (Lontok et al., 2015). Based on these observations, a revision of the genetic content covered in educational programs and the implementation of science popularization initiatives seem imperative.\n\nSome limitations of this study should be mentioned. This investigation did not attempt to address the differences in knowledge of genetics among groups classified by characteristics such as sex, ethnic group, age, family history of inherited diseases or level of education. Instead, this study was intended to be focused solely on a general undergraduate population not studying biology or medicine. Furthermore, more universities in different cities should be sampled to have a national perspective on students’ insights about genetics. Overall, these results provide a glimpse of the students’ standpoint toward genetics and its involvement in disease. Nevertheless, more effort is decisively needed to design and execute plans that will ensure an optimized method to measure knowledge of genetics in a larger and more diverse population. The data generated using these approaches will be proven essential when designing educational programs involving genetics and health. The application of such programs will be fundamental for the general population to avoid misunderstandings about genetics and to avoid the incorrect utilization of scientific terms.\n\nFollow-up studies will try to explore the knowledge about genetics and the attitudes toward related subjects, including genetic testing, stem cells, regenerative medicine and genetically modified organisms (GMOs). The expected results will provide improved insight into the population’s knowledge and will serve as a foundation for developing better strategies to increase the level of genetic literacy in our community.\n\n\nData availability\n\nOpen Science Framework: Assessing genetic knowledge in Ecuador. https://dx.doi.org/10.17605/OSF.IO/ZUVMN (Larrea, 2019)\n\nThis project contains the following extended data:\n\nDataset 1.pdf (the questionnaire in the original Spanish language and its translation into English)\n\nDataset 2.pdf (the MATLAB script to reproduce the analysis)\n\nOpen Science Framework: Assessing genetic knowledge in Ecuador. https://dx.doi.org/10.17605/OSF.IO/ZUVMN (Larrea, 2019)\n\nThis project contains the following underlying data:\n\nDataset 3.csv (a spreadsheet containing all responses to the evaluation)", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nThe authors want to thank the collaboration of UDLA Medicine Students of Molecular Biology (MDE-402: 2018-2, 2019-1) as part of a class exercise and discussion. Life Science Initiative supported the associated expenses, and Universidad de las Américas supported the paper processing fees.\n\n\nReferences\n\nBurton A: Are we ready for direct-to-consumer genetic testing? Lancet Neurol. 2015; 14(2): 138–39. PubMed Abstract \| Publisher Full Text\n\nCalsbeek H, Morren M, Bensing J, et al.: Knowledge and attitudes towards genetic testing: a two year follow-up study in patients with asthma, diabetes mellitus and cardiovascular disease. J Genet Couns. 2007; 16(4): 493–504. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCayon E, Correa J, Sarmiento-Sabogal J: Does Attending a Public or Private University Make a Difference for Students in Colombia? International Review of Management and Marketing. 2017; 7(2): 293–99. Reference Source\n\nCovolo L, Rubinelli S, Ceretti E, et al.: Internet-Based Direct-to-Consumer Genetic Testing: A Systematic Review. J Med Internet Res. 2015; 17(12): e279. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFitzgerald-Butt SM, Bodine A, Fry KM, et al.: Measuring genetic knowledge: a brief survey instrument for adolescents and adults. Clin Genet. 2016; 89(2): 235–43. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHaga SB, Barry WT, Mills R, et al.: Public knowledge of and attitudes toward genetics and genetic testing. Genet Test Mol Biomarkers. 2013; 17(4): 327–35. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHollands GJ, French DP, Griffin SJ, et al.: The impact of communicating genetic risks of disease on risk-reducing health behaviour: systematic review with meta-analysis. BMJ. 2016; 352: i1102. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHooker GW, Peay H, Erby L, et al.: Genetic literacy and patient perceptions of IBD testing utility and disease control: a randomized vignette study of genetic testing. Inflamm Bowel Dis. 2014; 20(5): 901–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJallinoja P, Aro AR: Knowledge about Genes and Heredity among Finns. New Genet Soc. 1999; 18(1): 101–10. Publisher Full Text\n\nLarrea M: Assessing genetic knowledge in Ecuador. OSF. 2019. https://www.doi.org/10.17605/OSF.IO/ZUVMN\n\nLea DH, Kaphingst KA, Bowen D, et al.: Communicating genetic and genomic information: health literacy and numeracy considerations. Public Health Genomics. 2011; 14(4–5): 279–89. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLontok KS, Zhang H, Dougherty MJ: Assessing the Genetics Content in the Next Generation Science Standards. PLoS One. 2015; 10(7): e0132742. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPerkins BA, Caskey CT, Brar P, et al.: Precision medicine screening using whole-genome sequencing and advanced imaging to identify disease risk in adults. Proc Natl Acad Sci U S A. 2018; 115(14): 3686–3691. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPhillips AM: 'Only a click away - DTC genetics for ancestry, health, love…and more: A view of the business and regulatory landscape'. Appl Transl Genom. 2016; 8: 16–22. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRafiq M, Ianuale C, Ricciardi W, et al.: Direct-to-consumer genetic testing: a systematic review of european guidelines, recommendations, and position statements. Genet Test Mol Biomarkers. 2015; 19(10): 535–47. PubMed Abstract \| Publisher Full Text\n\nRoberts J, Middleton A: Genetics in the 21st Century: Implications for patients, consumers and citizens [version 2; referees: 4 approved]. F1000Res. 2018; 6: 2020. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWu S, Pollard J, Chowdry A, et al.: Addressing the accuracy of direct-to-consumer genetic testing. Genet Med. 2019; 21(3): 758–759. PubMed Abstract \| Publisher Full Text" }	[ { "id": "47621", "date": "07 May 2019", "name": "Vasiliki Mollaki", "expertise": [ "Reviewer Expertise Human genetics", "molecular genetics", "bioethics" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe study aimed to assess the knowledge of genetics in a selected population of students from both public and private institutions.\nThe work is described in sufficient detail (regarding the samples size, study design and statistical analysis). The authors acknowledge the study limitations.\nAlthough similar studies have been performed in other countries, this is the first study assessing genetic knowledge in Ecuador. The result of the study could be proved useful for policy makers and educational programs.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "47624", "date": "21 May 2019", "name": "Rebecca Carver", "expertise": [ "Reviewer Expertise Science communication", "Public understanding of Genetics", "Science Education", "Public Health and Survey design" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study provides valuable insight into the level of knowledge of genetics in an Ecuadorian population.\nIntroduction Be careful with terminology regarding “perspectives” towards genetics, as this is not the same as knowledge about genetics. The same goes for “attitudes”.\nStudy design The first sentence in the methods section does not make sense. How is this a “baseline report on the attitude among undergraduate students toward genetic concepts?” What do the authors mean? Knowledge is not the same as attitudes.\nThe authors of the chosen instrument (Fitzgerald-Butt et al 20161) have stated in their paper that the instrument can be used for older teenage and young adult patients and parents in the pediatric setting, and that is most informative for individuals with below average genetic knowledge. How is this instrument applicable to the current study’s Ecuadorian student population, which is a very different sample population than the one for which the instrument was developed? What were the reasons for choosing this particular instrument? Can the authors justify their choice of instrument?\nResults What percentage of correct answers is “adequate” in the chosen instrument? Is there a reference value in the original instrument that can used to compare the results of this survey? (E.g. what minimum percentage of answers should the respondents answer correctly in order to have an adequate amount of knowledge?).\nDiscussion The results show that students achieved slightly better results when asked about genetics and diseases than when asked about genetic facts. One possible explanation for this could be that the questions about disease relate more to people’s lives than genetic facts. The authors state that they were surprised by the lower score obtained on a question about how chromosomes are passed to the next generation, but students may regard this type of “textbook knowledge” as less relevant to their everyday lives, and thus be less inclined to remember it. There is a lot of literature on how the genetics content in textbooks is lagging behind modern scientific developments – e.g. see previous literature by Gericke et al. and Dougherty et al. – and this may be an explanation for the results in this study. The authors might like to comment on this.\nConclusions The authors conclude by saying that despite the relatively adequate score overall (66.8% correct answers), the possibility that the correct responses were given by chance cannot be ignored – and that the actual knowledge of genetics among the participants might be less that that revealed by their answers. Consequently, the authors assert; “it seems evident that the national education programs of Ecuador require improvement in teaching of genetic concepts”. However, if the responses were given by chance, is it not also possible for the actual knowledge of genetics among the participants to be higher than that revealed here? Could the authors comment on this?\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-290
https://f1000research.com/articles/8-1467/v1	19 Aug 19	{ "type": "Case Report", "title": "Case Report: Spontaneous perforation of choledochal cyst in an infant: Successful management in a centre with limited means", "authors": [ "Rafey Abdul Rahman", "Umesh Kumar Gupta", "Rafey Abdul Rahman" ], "abstract": "Background: Biliary peritonitis due to a ruptured choledochal cyst (CC) is a rare occurrence. The difference between bile duct perforation (BDP) and ruptured choledochal cysts continues to be a matter of debate. Simple drainage, T tube placement and cholecystostomy have been proposed as the initial treatment of choice. Definitive surgery in the form excision of the CC and hepatico-enterostomy has been described as the ideal treatment option. We report a successful management of a unique case of perforated choledochal cyst in an infant who presented with biliary peritonitis. Case report: An 8 months old female child presented with biliary peritonitis as result of spontaneous perforation of a choledochal cyst. The patient was successfully managed initially by placement of T tube in the perforated cyst followed by a T tube cholangiogram. Definitive surgery was performed 5 weeks after the initial surgery in which cyst was excised and hepatico-duodenostomy was performed. The child is currently in follow up and doing well. Conclusion: Perforated CC can present as acute abdomen sometimes having only subtle signs. In absence of any previous established diagnosis of CC and trained radiological support the condition becomes challenging to diagnose preoperatively. External T-tube drainage followed by T-tube cholangiogram can help in delineating the anatomy. Cyst excision along with hepaticoportoentersomy remains the gold standard definitive treatment.", "keywords": [ "Spontaneous Choledochal cyst perforation", "bile duct perforation", "T tube cholangiogram", "hepatico-duodenostomy" ], "content": "Introduction\n\nBiliary peritonitis due to a ruptured choledochal cyst (CC) is a rare occurrence1,2. 15–25% of the cases of CC present with a classical triad of pain, lump and jaundice1. Cholangitis, pancreatitis, progressive biliary cirrhosis, stones in the cyst, portal hypertension and malignant transformation in the biliary tree are well known complications seen in CC3. Bile duct perforation and choledochal cysts are considered as inter-related pathology having a common cause in the form of pancreato-biliary mal-junction (PBM)4. The difference between bile duct perforation (BDP) and ruptured choledochal cysts continues to be a matter of debate. Simple drainage, T tube placement and cholecystostomy have been proposed as the initial treatment of choice. Definitive surgery in the form excision of the CC and hepatico-enterostomy has been described as the ideal treatment option. In this case report we describe the successful management of a case of a ruptured choledochal cyst in a female infant at a centre with limited means.\n\n\nCase report\n\nAn eight months old female child of South Asian ethnicity was admitted in the department of Paediatric medicine in April 2019 with complaints of progressive abdominal distension, non-passage of stools and mild generalised pain in abdomen for 3 days. There was no history of vomiting, fever or jaundice. On clinical examination features of peritonitis were observed. A rectal examination was performed and was found to be insignificant. An X Ray of the abdomen in the erect position was performed, with results suggestive of ascites with paucity of air in the pelvis. An ultrasound of the abdomen revealed free fluid in the abdomen, along with mildly dilated intrahepatic biliary radicles. The patient’s complete blood work up revealed low levels of haemoglobin at 9.8 grams (Normal range: 12–14 gm/dl), while the rest of the parameters were within the normal limits. In the liver function tests (LFT), serum glutamic oxaloacetic transaminase (SGOT) and SGPT were mildly elevated with values of 140 and 143 units/litre respectively. The remaining LFT parameters were within the normal range. As peritonitis was suspected an emergency laparotomy was performed. Biliary peritonitis secondary to a perforation in the common bile duct (CBD) was observed and the CBD was grossly dilated. The wall of the small and large bowel was oedematous but otherwise appeared normal. Bile stained ascitic fluid was drained and a “T tube” was placed through the perforation site in the CBD. (Figure 1). Post operatively bile output through the “T Tube” was 150–200 ml per day which was high. We expected it to be around 50 ml per day thinking that some amount of bile may pass into the gut through the “T tube”. However, as the system was obstructed the output through the “T tube” remained persistently high. The child was transfused with packed red blood cells and fresh frozen plasma post-operatively. Other supportive treatment in the form of intravenous (IV) fluid, antibiotics and analgesics were administered. IV antibiotics included Cefotaxime (90 mg/kg/day in 3 divided doses), Amikacin (15mg/kg/day once a day) and Metronidazole (20mg/kg/day in 3 divided doses) covering gram positive, gram negative and anaerobic bacteria. The child was started on oral feeds from post-operative day 5. The child was discharged on postoperative day 8 with the “T tube” but was readmitted again with features of dehydration. The patient was stabilised again and a “T Tube” cholangiogram was performed in the third week post-surgery (Figure 2) which revealed a dilated CC along with dilated intrahepatic biliary radicles with no contrast flow into the duodenum (Figure 3). We waited for another 2 weeks and built up the child nutritionally for the definitive surgery. In view of the absence of a trained radiologist and a magnetic resonance imaging (MRI) scanner at the institute, Magnetic Resonance Cholangiopancreatography (MRCP) could not be done. Hence, we decided to proceed with definitive surgery based on the “T Tube” cholangiogram imaging and previous operative findings. A laparotomy was performed. The dilated cyst was excised completely after separating it from the portal vein. The distal end of the cyst was blind ending and was not communicating with the duodenum. Distally the cyst was excised at its narrowest portion on the duodenum. Proximally, common hepatic duct was identified and then a hepatico-duodenostomy was performed. The child recovered well and started passing cholic stools from post-operative day 3. The child was started on oral feeds from post-operative day 5. The patient was discharged on post-operative day 8 on full oral feeds. The patient is currently in our follow up and doing well.\n\n\nDiscussion\n\nCC is considered to be a congenital anomaly as it seen in foetuses and newborns. Its etiology is unknown, however many theories exist including abnormalities in the pancreatico-biliary junction (PBJ)4, distal obstruction with congenital weakness of the duct5 and ischemia of the bile duct6. CC rupture and resulting peritonitis is a rare complication, reported in only 1.8% to 2.8% of cases2.\n\nUltrasonography is a useful diagnostic investigation for CC when the cyst is intact, but in cases of perforation of the cyst, where the cyst decompresses, free fluid in the abdomen may be the only finding. Pre-operative hepato-biliary scintigraphy can demonstrate bile leak from the extra hepatic biliary tree in biliary channel perforation7. MRCP is a non-invasive method of evaluating the biliary tree, and owing to its excellent anatomical and contrast resolution, it can be used to diagnose bile duct perforation, perforated CC and PBM8. The cost and duration of MRCP imaging limits its use.\n\nOur case presented with features of peritonitis and USG demonstrated free fluid in the abdomen. In the absence of a previous diagnosis and no past history suggestive of choledochal cyst, it was difficult for us to accurately diagnose the condition preoperatively. Since the centre had no facility for hepatobiliary scintigraphy and MRCP, or a trained radiologist, we went ahead with laparotomy and were able to diagnose and treat the patient with external biliary diversion. Since there was no precipitating factor for perforation of CC like trauma, we presumed the rupture was spontaneous.\n\nThrough our intraoperative findings and the “T tube” cholangiogram we were able to identify the CC and perforation site. Absence of a trained radiologist and with no close alternative radiological centre, as our institution is situated in a rural area, we were unable to perform additional diagnostic imaging like MRCP. In the post-operative week 5 we went ahead with definitive surgery. We performed complete cyst excision and a hepatico-duodenostomy. We chose hepatico-duodenostomy over hepatico-jejunostomy for various reasons. Firstly, it reduced the duration of surgery; secondly, the duodenum was mobile reaching easily up to the common hepatic duct; and thirdly the literature shows similar success rates and complications for both procedures9.\n\nIt was also important for us to distinguish BDP from a ruptured CC as management of both differs. BDP occurs exclusively in infants less than 20 weeks of age and perforation mostly occurs at the junction of cystic duct and common hepatic duct2. Simple surgical drainage often leads to closure of BDP but a perforated CC requires cyst excision and heptico-enterostomy.\n\n\nConclusion\n\nPerforated CC can present as biliary peritonitis sometimes having only subtle signs. In absence of any previous established diagnosis of CC and trained radiological support becomes challenging to diagnose this preoperatively. External T-tube drainage followed by T-tube cholangiogram can help in revealing the anatomy of the hepatobiliary tree. Cyst excision and hepatico-enterostomy remains the gold standard definitive treatment.\n\n\nConsent\n\nWritten informed consent for publication of the patients’ clinical details and clinical images was obtained from the parents of the patient.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nChen W, Chang C, Hung W: Congenital choledochal cyst: with observations on rupture of the cyst and intrahepatic ductal dilatation. J Pediatr Surg. 1973; 8(4): 529–538. PubMed Abstract \| Publisher Full Text\n\nTreem WR, Hyams JS, McGowan GS, et al.: Spontaneous rupture of a choledochal cyst: clues to diagnosis and etiology. J Pediatr Gastroenterol Nutr. 1991; 13(3): 301–306. PubMed Abstract\n\nChijiiva K, Koga A: Surgical management and long-term follow-up of patients with choledochal cysts. Am J Surg. 1993; 165(2): 238–242. PubMed Abstract \| Publisher Full Text\n\nOhkawa H, Takahashi H, Maie M: A malformation of the pancreatico-biliary system as a cause of perforation of the biliary tract in childhood. J Pediatr Surg. 1977; 12(4): 541–6. PubMed Abstract \| Publisher Full Text\n\nSuda K, Matsumoto Y, Miyano T: Narrow duct segment distal to choledochal cyst. Am J Gastroenterol. 1991; 86(9): 1259–1263. PubMed Abstract\n\nLloyd DA, Mickel RE: Spontaneous perforation of the extra-hepatic bile ducts in neonates and infants. Br J Surg. 1980; 67(9): 621–623. PubMed Abstract \| Publisher Full Text\n\nPaladugu R, Rau A, Schein M, et al.: Spontaneous perforation of the hepatic duct in adults. Dig Surg. 1998; 15(5): 417–20. PubMed Abstract \| Publisher Full Text\n\nSugiyama M, Baba M, Atomi Y, et al.: Diagnosis of anomalous pancreaticobiliary junction: value of magnetic resonance cholangiopancreatography. Surgery. 1998; 123(4): 391–7. PubMed Abstract \| Publisher Full Text\n\nNarayanan SK, Chen Y, Narasimhan KL, et al.: Hepaticoduodenostomy versus hepaticojejunostomy after resection of choledochal cyst: A systematic review and meta-analysis. J Pediatr Surg. 2013; 48(11): 2336–2342. PubMed Abstract \| Publisher Full Text" }	[ { "id": "55411", "date": "22 Oct 2019", "name": "Anand Pandey", "expertise": [ "Reviewer Expertise Pediatric Surgery" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIt is an unusual report where patient was managed successfully.\n\nAttention is needed towards grammatical errors, which have crept in at places.\n\nSpecific values of bilirubin are missing. Was there any history of clay coloured stools in retrospective assessment?\n\nThere is repetition in second paragraph page 4 about MRCP. It may be deleted.\n\nLiterature review may include other articles on ruptured choledochal cyst such as - Management of rupture of choledochal cyst. Ahmed I et al. Indian J Gastroenterol. 2011 Mar;30(2):94-61.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Partly\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Partly", "responses": [ { "c_id": "5008", "date": "11 Nov 2019", "name": "umesh gupta", "role": "Author Response", "response": "Specific values of bilirubin are missing.Specific values of bilirubin was total 0.8 mg %.Was there any history of clay coloured stools in retrospective assessment?No clay coloured stools was present in retrospective assessment ." } ] }, { "id": "59758", "date": "17 Feb 2020", "name": "Richard Kozarek", "expertise": [ "Reviewer Expertise Areas of expertise range from therapeutic endoscopy", "inflammatory bowel diseases", "and practice economics" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe case report by Rahman and Gupta describing an 8-month-old female with spontaneous perforation of a congenital choledochal cyst (CC) demonstrates the ability to provide appropriate patient care despite a paucity of diagnostic options. These providers had only an abdominal ultrasound which demonstrated intra-abdominal fluid following spontaneous perforation of a CC. Surgical exploration revealed bile peritonitis and a local bile duct perforation treated with a T-tube decompression. Diagnosis as a CC was only made at subsequent cholangiography leading appropriately to cyst resection and choledochal-enteric biliary anastomosis. All this was done without access to MRI/MRCP, apparently a CT scan, or access to an onsite radiologist. As a gastroenterologist who has provided ERCP support to pediatric patients for > 30 years, one who has been involved in the treatment of more than 100 adult patients with CC, and an individual who has been involved in education and training programs in the developing world for a comparable time period, I salute their diagnostic tenacity, appropriate treatment, and apparent outcomes.\n\nIn areas in which more resources are available, it is unlikely that this patient would have been explored without a CT scan, and if a grossly dilated common bile duct had been defined, a subsequent MRCP to define the extent of the cyst as used the pancreaticobiliary junction to define the type and extent of surgical resection. Although the authors suggest that the patient is doing well after CC resection and hepaticoduodenostomy, there is no report of the follow-up times. A follow-up 2 weeks after surgery is much less important than a 1 year follow-up with normal liver enzymes and growth parameters. The follow-up interval needs to be defined.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? No\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1467
https://f1000research.com/articles/8-1466/v1	19 Aug 19	{ "type": "Research Article", "title": "Growth performance and survival rate of Portunus pelagicus (Linnaeus, 1758) broodstock females fed varying doses of amaranth extracts", "authors": [ "Efrizal Efrizal", "Zuhri Syam", "Rusnam Rusnam", "Suryati Suryati", "Zuhri Syam", "Rusnam Rusnam", "Suryati Suryati" ], "abstract": "Background: Formulated diets made from food waste enriched with amaranth extract largely determine the level of softshell crab production. The study was carried out to analyze the growth performance and survival rate of blue swimming crab Portunus pelagicus (Linnaeus, 1758) broodstock females fed formulated diets with different doses of amaranth extracts. Methods: The female crab samples were collected from the coastal region of Padang, West Sumatra. The method used in this study was a completely randomized design, with four treatments and five replications of amaranth extract-enriched diets (0, 250, 500, and 750 ng/g crab). Results: The results show that the enrichment of the formulated diet with amaranth extracts significantly affected (P < 0.05) the absolute weight gain (AWG), carapace length (ACL), and carapace width (ACW). Conclusions: A quadratic relationship existed between the amaranth extract dose in the formulated diet and the AWG, ACL and ACW.", "keywords": [ "Blue swimming crab", "broodstock", "formulated diets", "growth performance", "amaranth extracts" ], "content": "Introduction\n\nThe waters of the province of West Sumatra provide a water marine area that includes the Indonesian Exclusive Economic Zone, which encompasses approximately 186,580 km2, with a long coastline of 2420.387 km, and provides habitat for many biological natural resources1. This marine area, when used wisely, from planning to the administration, implementation, and supervision, will support the welfare of people along the coast in general and the coast of Indonesia in particular. The crab (Portunus pelagicus) belongs to the Portunidae family, is a marine biota resource. It has great potential to become an important fishery export commodity2–6, since in recent years, domestic and international demand has increased from year to year7,8. According to data from the last ten years for the period 1993–2002, the export volume of crabs increased by an average of 16.72% per year, from 6,081 tons in 1993 to 11,246 tons in 20021, whereas in the period 2007–2009 reports on the Indonesian market share of crab decreased from 17.6% to 16.3%9. The intensive harvest of crabs could cause a decline in natural populations of crabs10. Because little control exists in crab harvesting, populations of crabs are rare in Indonesian waters11. Until now the demand for crabs, for domestic consumption and for export, has still relied on the catch from the sea, so this concern will affect the crab population in nature12.\n\nAn increase in P. pelagicus production can be achieved in an intensive aquaculture system. Aquaculture is one of the ways to address the threat of crab population decline from overexploitation13–16. The business of crab farming is still in its early phase in Indonesia and in many countries in the world17. Some researchers have reported that a low growth performance and the survival rate remain as problems in culture of the blue swimming crab17, which is caused by various factors, such as disease18–20, molting syndrome21, cannibalism22, and feed3,7,8. Feed is the main component needed by the crab to survive and grow. The completeness of nutrients in the feed is absolutely necessary to ensure the normal growth of the crabs. Nutritional requirements for crab growth, including proteins, fats, carbohydrates, vitamins, and minerals, differ by crab type and size3,8,23, with advantages being provided from formulated feeds over the live feeds (fresh feed)24.\n\nOne important breakthrough in the cultivation of softshell crabs, which has been developed by Fujaya25, is the discovery of a molting stimulant derived from amaranth plant extract (Amaranthacea tricolor) called vitomolt. Amaranth extract containing ecdysteroids and application through injection proved to accelerate molting and did not cause death in the crabs; furthermore, the growth of crabs that received the application of amaranth extract was greater than that of the crabs that did not. Based on the description given above, an in-depth assessment of the effects of the feeding of amaranth extracts in varying doses on the quality of formulated diets for growth performance and survival rate of P. pelagicus (Linnaeus, 1758) was needed. Therefore, this study was conducted.\n\n\nMethods\n\nThe study was conducted from January until June 2019 at Balai Benih Ikan Pantai (BBIP) Teluk Buo, Balai Benih Ikan (BBI) Bungus, in the city of Padang and at the Laboratory of Animal Physiology Department of Biology, Padang, West Sumatra.\n\nIn recent years, the production of crabs has decreased both in number and size. While Blue Swimming crab (Portunus pelagicus) is classified as vulnerable to endangered in the Indonesian marine waters, the Government of the Republic of Indonesia does not require licenses to be obtain to capture and rear this species (Regulation of the Minister of Environment and Forestry of the Republic of Indonesia No: P.92/MENLHK/SETJEN/KUM.1/8/2018; https://www.jalaksuren.net/wp-content/uploads/2018/10/Permen-LHK-P.92-2018.pdf), hence no licenses are applicable to this study. No animals suffered as a result of the activities of this study. P. pelagicus was transported to the hatchery for rearing and at the end of the experiment P. pelagicus remained in good condition.\n\nIn total 70 female crab samples were collected on January 2019; 20 were females at stage II of ovarian maturity, which were selected for this study2. The female crab samples were collected from the coastal region of Padang (100° 22’12” BT, 1° 0’ 54” LS dan 100° 25’58,8” BT, 1° 4’ 40,8” LS), West Sumatra, and the crabs were placed randomly in four concrete tanks (200 cm × 100 cm × 100 cm), each divided into five units, i.e., plastic boxes (45.5 cm × 32.5 cm × 16.5 cm), where the crabs were held at a maximum density of one crab per box. Tanks were provided with approximately a 15-cm-thick layer of sand for substrate and with adequate aeration2,7,8,26–28. The crabs were maintained in monitored water with a water depth of 25 to 30 cm, salinity of 30 to 32 ppt, pH 7.77 to 7.96, temperature 26 to 27°C, and dissolved oxygen (DO) 7.00 to 7.30 ppm. Each crab was provided with a shelter made of PVC pipe, 13 cm in diameter and 40 cm in length, to serve as a refuge during molting. Dietary Vitamin E was fed daily at 3% biomass (17.00 to 18.00 hours), and uneaten food was removed every morning.\n\nThe method used in this study was a completely randomized design (CRD) with four treatments and five replications of amaranth extract-enriched diets, with treatments as follows: Fdiet 1, formulated diet without amaranth extract 0 ng/g crab; Fdiet 2, formulated diet enriched with amaranth extract at 250 ng/g crab; Fdiet 3, formulated diet enriched with amaranth extract at 500 ng/g crab; and Fdiet 4, formulated diet enriched with amaranth extract at 750 ng/g crab. The formulated diet3,7,29 was a modified formulation of one used for the broodstock of mud crab Scylla serrata16. Amaranth extract was dissolved in 80% ethanol in a ratio of 1:1 and then homogenized. Then, the 80% ethanol solution was added at a rate of 20 mL/kg of feed by even spraying of the solution onto the test feed, and the feed was then left to dry. The test feed was stored until it was ready for use30.\n\nPrior to receiving artificial feed, the broodstock was first given a natural feed (fresh bivalve molluska + sardinella fish; 1: 1) and gradually acclimatized with artificial feed for 10 days. Feed was given at a dose of 8–10% of biomass per day for natural feed and 3–4% for artificial feed. The feed was given three times a day at 08.00, 13.00 and 17.00, with a percentage of 40% in the morning and the remainder being divided into two periods: the afternoon and evening. The remaining food was discarded every morning, and the amount of food was adjusted to the weight of the crab mother at the time of observation. The number of dead crabs were observed and recorded daily.\n\nMeasured parameters (absolute weight gain, absolute carapace length, absolute carapace width and survival rate) and water quality refer to Efrizal et al.7. The absolute weight gain was calculated as follows: AWG = WGf – WGo, where AWG is weight gain (g), WGf is the weight of the crab at the end of experiment (g), and WGo is the weight of crab at the start of experiment (g). The absolute carapace length was calculated as follows: ACL = CLf – CLo, where ACL is carapace length gain (mm), CLf is the carapace length of the crab at the end of experiment (mm), and CLo is the carapace length of crab at the start of experiment (mm). The absolute carapace width was calculated as follows: ACW = CWf – CWo, where ACW is carapace width gain (mm), CWf is the carapace width of the crab at the end of experiment (mm), and CWo is the carapace width of crab at the start of experiment (mm). The carapace length, carapace width, and survival rates were measured as described previously2,3,31. Weights were measured to 0.01 g on the electronic balances (BL3200H-SHIMADZU).\n\nThe water quality parameters that were monitored daily were temperature (°C), salinity (ppt), pH, and water depth (cm) while dissolved oxygen (ppm) and CO2 (ppm) thrice weekly using a maximum-minimum thermometer, hand-held Atago refractometer model 8808, Thermo Orion Benctop pH meter models 410 A plus, weighted line, YSI oxygen meter model 57, and APHA32, respectively.\n\nThe data for growth performance (absolute weight gain, absolute carapace length, and absolute carapace width) and survival rate7 were analyzed using one-way ANOVA and Duncan’s Multiple Range test to compare the differences among the means of the different treatments33 were performed using SPSS software (version 19.0 for Windows; SPSS Inc., Chicago, II. Arcsine transformation was performed before the data, in percentages, was analyzed.\n\n\nResults\n\nThe growth measure of absolute weight is a measure of the weight difference observed in a female crab for a certain time when weighed at the beginning and end of the period. The observation of the average weight and absolute weight gain in the female parent crab P. pelagicus (Linnaeus, 1758) with different dietary treatments is presented in Table 1. Table 1 shows that the growth tends to increase the absolute weight of the maintenance period for 0 to 40 days. From the weighing results (Table 1), the highest absolute value of the average weight obtained in the administration was achieved with Fdiet 3 (55.03 g), followed by Fdiet 4 (18.05 g), Fdiet 1 (32.49 g), and Fdiet 2 (42.49 g); analysis of variance showed significant differences (P < 0.05). Duncan's test showed further significant differences (P < 0.05), which can be observed between the treatments Fdiet 3 and Fdiet 1 and between Fdiet 3 and Fdiet 4, whereas no significant differences (P > 0.05) were observed between Fdiet 3 and Fdiet 2, and between Fdiet 1 and Fdiet 4. The relationship between the amaranth-extract doses in the formulated diet and the absolute weight gain is presented in Figure 1. The regression equation is AWG = -0.0002DAE + 0.1274DAE + 29.816 (R² = 0.4155; P < 0.05).\n\nValues are means ± standard errors (SE). AWG, absolute weight gain; Fdiet 1, formulated diet without amaranth extract 0 ng/g crab; Fdiet 2, formulated diet enriched with amaranth extract 250 ng/g crab; Fdiet 3, formulated diet enriched with amaranth extract 500 ng/g crab; and Fdiet 4, formulated diet enriched with amaranth extract 750 ng/g crab.\n\nAbsolute carapace length is calculated from the difference in the size of the parent crab carapace length achieved within a certain time, when the carapace length is measured at the beginning and end of the period. Treatment with the different diets caused relatively large changes in the growth of the absolute carapace length during the maintenance period for the 40 days, ranging from 1.22 to 7.09 mm (Table 2), with analysis of variance showing significant differences (P < 0.05). Based on the data in Table 2, the accretion of the greatest absolute carapace length was obtained with the treatment Fdiet 3 (7.09 mm) in comparison with the treatment Fdiet 4 (1.22 mm), Fdiet 1 (3.60 mm) and Fdiet 2 (4.61 mm). Similarly, the results of a further test for significant differences (P < 0.05) with Duncan’s Test showing that treatment Fdiet 3 differs from Fdiet 4, whereas the treatment Fdiet 3 does differ significantly from Fdiet 1 and Fdiet 2 (P < 0.05), and Fdiet 4 is not significantly different from Fdiet 1 and Fdiet 2 (P > 0.05). The relationship between the dose of the amaranth extracts in the formulated diet and absolute carapace length (mm) of the blue swimming crab P. pelagicus (Linnaeus, 1758) female broodstock was quadratic (ACL = -2 x 10-5DAE2+ 0.0166DAE+3.1695; R² = 0.2212; P < 0.05; Figure 2).\n\nValues are means ± standard errors (SE). ACL, absolute carapace length; Fdiet 1, formulated diet without amaranth extract 0 ng/g crab; Fdiet 2, formulated diet enriched with amaranth extract 250 ng/g crab; Fdiet 3, formulated diet enriched with amaranth extract 500 ng/g crab; and Fdiet 4, formulated diet enriched with amaranth extract 750 ng/g crab.\n\nThe growth of the absolute carapace width is also calculated from the difference in the size of the parent crab carapace width achieved in a specific time, when the carapace width is measured at the beginning and end of the period. From the measurement results (Table 3), artificial feeding at Fdiet 3, a dose of 500 ng of amaranth extracts/g crab provides relatively high added value to the mean carapace width (13.14 mm) compared to artificial feeding at Fdiet 1, an amaranth extract dose of 0 ng/g crab (6.25 mm); Fdiet 2, 250 ng/g crab (8.98 mm); and Fdiet 3, 750 ng/g crab (1.92 mm), with analysis of variance showing a significant difference (P < 0.05). Similarly, Duncan's test results showed significant differences (P < 0.05), as seen between the treatment Fdiet 3 with Fdiet 1 and Fdiet 4, whereas the treatment Fdiet 2 with Fdiet 3 and the treatment Fdiet 1 with Fdiet 4 did not show significant differences (P > 0.05). The relationship between the amaranth-extract dose in the formulated diet and absolute carapace width was found to be quadratic (ACW = -5 x 10-5 DAE2- 0.0349DAE+5.502; R² = 0.1724; P > 0.05; Figure 3).\n\nValues are means ± standard errors (SE). ACW, absolute carapace width; Fdiet 1, formulated diet without amaranth extract 0 ng/g crab; Fdiet 2, formulated diet enriched with amaranth extract 250 ng/g crab; Fdiet 3, formulated diet enriched with amaranth extract 500 ng/g crab; and Fdiet 4, formulated diet enriched with amaranth extract 750 ng/g crab.\n\nThe results show that the different dietary treatments for the female parent crabs during the maintenance period of 40 days in controlled cultivation containers provided a high survival rate (100%) for all treatments (Table 4). The high survival rate value is due to the maintenance in a controlled container cultivation, where no deaths occurred in the female parent crabs. This happens because the water quality (physical and chemical factors) during the study remained in the range conducive for crabs (Table 5).\n\nValues are means ± standard errors (SE). H, survival rate; Fdiet 1, formulated diet without amaranth extract 0 ng/g crab; Fdiet 2, formulated diet enriched with amaranth extract 250 ng/g crab; Fdiet 3, formulated diet enriched with amaranth extract 500 ng/g crab; and Fdiet 4, formulated diet enriched with amaranth extract 750 ng/g crab.\n\n\nDiscussion\n\nThe growth responses of the crab to the four formulated diets with amaranth extract occurred because of the varied composition of the raw materials used in formulations. Our results demonstrate that the different nutrient levels in the feed, which are especially caused by the levels of amaranth extract. The blue swimming crab needs to maintain the existence of life and its growth, and it will grow well if the available feed contains all the nutrients needed at optimal levels. According to Gutierrez-Yurrita and Montes34, the nutrient composition of essential feed will determine the growth and efficiency of the organisms being fed.\n\nIn this experiment (Table 1), the formulated diet that included a dose of amaranth extract influenced the physiological processes of the experimental crab known as absolute weight gain. The highest weight gain was shown by the test crabs receiving the Fdiet 3 treatment (55.03 ± 2.06 g) compared to other feed treatments. The absolute weight gain decreased dramatically with an increase in the amaranth extract dose of 750 ng/g crab (18.05 ± 8.27 g). This dose (Fdiet 3, 500 ng/g crab) performed well, giving good results. In contrast to the findings of the Fdiet 3 treatment, a decrease occurred in the absolute weight of crabs receiving the Fdiet 4 treatment, apparently because of a treatment overdose effect, which causes hormonal imbalances in the body, thereby affecting the absolute weight gain.\n\nAmaranth extract is a molting stimulant containing molting hormone (phytoecdysteroid). Ecdysteroids are the main steroid hormone in arthropods, which functions as a molting hormone and regulate physiological functions, such as growth, metamorphosis and reproduction35. This hormone is secreted by Y-organs in the form of ecdysone. The hemolymph then converts ecdysone into a 20-hydroxyecdysone (20E) active hormone, which is the enzyme 20-hydroxylase found in the epidermis of other organs and tissues. The titers of 20-hydroxyecdysone (20E) in circulation vary throughout the molting phase. Shortly after the ecdysis (molting), the titers are very low, where they remain throughout the intermolt phase.\n\nOur results cast a new light on the administration of amaranth extract at 0-500 ng/g crab, which caused an increase in absolute carapace length (3.60 ± 1.64 - 7.09 ± 0.12 mm) and contributed to relatively large absolute carapace widths (6.25 ± 2.97- 13.14 ± 0.65 mm); at higher doses (750 mg/g crab), a drastic decrease occurred, with successive values of 1.22 ± 1.37 and 1.92 ± 2.15 mm. The increase in absolute carapace length and absolute carapace width was apparently attributable to the synergistic cooperation of the hormones contained in amaranth extract with natural hormones in the body of the crab. This is supported by Fujaya et al.17, who reported that amaranth plants (Amaranthaceae tricolor) contain phytoecdysteroid, which acts as a stimulant for molting and the production of softshell crabs. Kuballa et al.36 noted that molting is physiologically controlled by the molting hormone. Wahyuningsih37 reported an increase in molting percentage by 54% in crabs that received vitomolt supplementation (phytoecdysteroid) at 250 ng/g crabs compared to controls that only showed a 15% increase. Meanwhile, in the study by Susanti38, a molting percentage of 90% was obtained after the vitomolt treatment in formulated diet (933 ng/g feed) compared to only 20%. A dramatic decrease at the higher doses in this study (Fdiet 4) is attributed to the threshold of phytoecdysteroid hormone levels in the blood that affects the molting process in the crab. In addition, a decrease in protein synthesis as a result of the disruption of the physiology of the molting hormone produced was also suspected. Techa and Chung39 suggested the most prominent metabolic action of steroids is activated protein metabolism. Preston and Dinan40 suggest that ecdysteroids also play a role in increasing protein formation by increasing the mRNA synthesis. According to Lafont and Dinan41, ecdysteroid also stimulates carbohydrate metabolism and lipid biosynthesis and acts as an immunostimulant and antioxidant.\n\nIt is important to note that the water quality is critical to the survival, health and growth of crabs, especially in semi-intensive and intensive culture. The physical and chemical properties of water should be kept within certain levels. According to Habashy and Hassan42, the required water quality for the maintenance of crustaceans includes salinity, temperature and pH43. Moreover, salinity affects other physiological processes of crustaceans, and various indices exist to evaluate their physiological responses and metabolism, which include glucose, superoxide dismutase and acid phosphatase44,45. Table 5 shows that water salinity between 30.0 to 32.0 ppt was observed in the present study. Salinity is therefore within the range that is highly favorable to the survival of the crab46–48. Mud crabs are highly tolerant to varying salinity conditions ranging between salinity of 10 ppt to 34 ppt49. Water temperature and changes in water temperatures have considerable influence on the rate of growth and survival of aquatic organisms. In this study, temperature during the observation ranged from 26.0 to 27.0°C; therefore, the temperature was in the range of support for the activities of life, growth and reproduction of crab50,51. The pH is a variable that is known to affect survival and development in many brachyuran species. In this study, the range of pH for survival was 7.77–7.96. The optimal pH for growth is from 8–8.552. Low pH can stress crustaceans and cause soft shell and poor survival. At a pH between 7.26 and 8.00, mortality is not usually observed7.\n\nMany studies have been conducted on the oxygen requirements of crustaceans52–54. Liao and Murai53 reported that the oxygen respiration of P. monodon remained constant at dissolved oxygen levels above 3.00 to 4.00 ppm at salinity 5–45 ppt and temperature 20-30°C. In the present study, dissolved oxygen was relatively high, ranging from 7.00–7.30 ppm. In heavily stocked crab ponds, the carbon dioxide can become high as a result of respiration. Boyd and Tucker55 explain that free carbon dioxide is good for the crustacean at no more than 12 ppm and must not be less than 2 ppm. In the present experiments, carbon dioxide levels ranged from 3.85–6.30 ppm (Table 5).\n\n\nConclusion\n\nThe present study demonstrated that the enrichment of formulated diet with amaranth extracts had a significant effect on the absolute weight gain (AWG), carapace length (ACL), and carapace width (ACW) of Portunus pelagicus (Linnaeus, 1758) broodstock females. The enrichment of amaranth extracts in formulated diets causes an increase in the AWG (18.05-55.03 g), ACL (1.22-7.09 mm) and ACW (1.92-13.14 mm) and yielded a 100% survival rate for all treatments during the 40-day maintenance period. Based on regression analysis, a quadratic relationship was observed between the dose of amaranth extracts found in the formulated diets and the AWG, ACL and ACW.\n\n\nData availability\n\nAll data underlying the results are available as part of the article (Table 1–Table 5) and no additional source data are required.", "appendix": "Grant information\n\nThis study was supported by a research grant from the Faculty of Math and Science, Andalas University (No. 17/UN.16.03.D/PP/FMIPA/2019) and “Percepatan Guru Besar” from the Research and Community Service Institution, Andalas University (No. T/49./UN.16.17/PP.OK-KRP2GB/LPPM/2019).\n\nThe funders had no role in study design, data collection and analysis, decisions to publish, or preparations of the manuscript.\n\n\nAcknowledgements\n\nThe authors wish to thank all the staff at Fish Seed Center Beaches (BBIP) Gulf Buo and Fish Seed Center (BBI) Bungus, the City of Padang, Department of Maritime and Fisheries Affairs, and the Laboratory of Animal Physiology Department of Biology, Faculty of Math and Science, Andalas University, Padang, West Sumatra for technical assistance rendered.\n\n\nReferences\n\nEfrizal: Kebijakan pemanfaatan wilayah sumberdaya pesisisr dan pantai Sumatera Barat. Prosiding Seminar dan Lokakarya Kelautan Regional Sumatera Ketiga Padang, 23 November 2000. Pusat Penelitian dan Pengembangan Perikanan. Fakultas Perikanan Universitas Bung Hatta, Padang. 2000.\n\nEfrizal E: Effects of stocking density on survival rate and larval development of blue swimming crab, Portunus pelagicus (Linnaeus, 1758) under laboratory conditions. AACL Bioflux. 2017a; 10(2): 217–226. Reference Source\n\nEfrizal, Rusnam: The Influence of Dietary Vitamin E on the Gonad Development of Blue Swimming Crab, Portunus pelagicus (Linnaeus, 1758). Aust J Basic Appl Sci. 2017b; 11(1): 7–15.\n\nFahmi AS, Maksum M, Suwondo E: USFDA Import Refusal and Export Competitiveness of Indonesian Crab in US Market. Agriculture and Agricultural Science Procedia. 2015; 3: 226–230. Publisher Full Text\n\nFAO: Fisheries and aquaculture software. FishStatJ - software for fishery statistical time series. In: FAO Fisheries and Aquaculture Department [online]. Rome. Updated 21 July 2016. [Cited 23 May 2018]. 2016. Reference Source\n\nKangas MI: Synopsis of the biology and exploitation of the blue swimmer crab, Portunus pelagicus Linnaeus, in Western Australia. Fisheries Research Report No. 121, Department of Fisheries, Western Australia, 2000; 22. Reference Source\n\nEfrizal, Rusnam, Suryati, et al.: Evaluation of Formulated Diets Enriched by Spinach Extracts for the Broodstock Females, Portunus pelagicus (Linnaeus, 1758). Pak J Biol Sci. 2019a; 22(6): 283–290. Publisher Full Text\n\nEfrizal E, Zakaria IJ, Rusnam R, et al.: Studies on Biological Test of Formulated Diets Supplementation of Vitamin E for the Broodstock of Females Blue Swimming Crab, Portunus pelagicus (Linnaeus, 1758) [version 2; peer review: 2 approved, 1 approved with reservations]. F1000Res. 2019; 7: 1780. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNatalia D, Nurozy: Performance of Indonesian fisheries products competitiveness in global market. Bulletin of Scientific Research and Development of Trade. 2012; 6(1). (in Bahasa Indonesia).\n\nLorentzen G, Voldnes G, Whitaker RD, et al.: Current Status of the Red King Crab (Paralithodes camtchaticus) and Snow Crab (Chionoecetes opilio) Industries in Norway. Reviews in Fisheries Science & Aquaculture. 2018; 26(1): 42–54. Publisher Full Text\n\nSara L, Muskita WH, Astuti O, et al.: Some population parameters of blue swimming crab (Portunus pelagicus) in Southeast Sulawesi waters, Indonesia. AACL Bioflux. 2017; 10(3): 587–601. Reference Source\n\nJuwana S: Crab culture technique at RDCO-LIPI, Jakarta, Indonesia 1994 to 2001. Proceeding Workshop on Mariculture in Indonesia. Mataram, Lombok Island. Research Center for Oceanography-LIPI, Institute of Marine Research Norwegian Bergen-Norway. 2002; 144.\n\nHarris D, Johnston D, Sporer E, et al.: Status of the blue swimmer crab fishery in Shark Bay, Western Australia. Fisheries Research Report No. 233, Department of Fisheries, Western Australia, 2014; 79. Reference Source\n\nJohnston D, Harris D, Caputi N, et al.: Decline of a blue swimmer crab (Portunus pelagicus) fishery in Western Australia - history, contributing factors and future management strategy. Fish Res. 2011; 109(1): 119–130. Publisher Full Text\n\nKunsook C, Gajaseni N, Paphavasit N: A Stock Assessment of the Blue Swimming Crab Portunus pelagicus (Linnaeus, 1758) for Sustainable Management in Kung Krabaen Bay, Gulf of Thailand. Trop Life Sci Res. 2014; 25(1): 41–59. PubMed Abstract \| Free Full Text\n\nMehanna SF, Khvorov S, Al-Sinawy M, et al.: Stock assessment of the blue swimmer crab Portunus pelagicus (Linnaeus, 1766) from the Oman Coastal Waters. Int J Fish Aquat Sci. 2013; 2(1): 1–8. Reference Source\n\nFujaya Y, Trijuno DD, Aslamyah S, et al.: Domestication and selective breeding for producing fast growing and high meat quality of blue swimming crab (Portunus pelagicus). AACL Bioflux. 2016; 9(3): 670–679. Reference Source\n\nGovindasamy C, Srinivasan R: Association of bioluminescent bacteria from blue swimmer crab Portunus pelagicus (Linneaus, 1758). Asian Pacific Journal of Tropical Disease. 2012; 2(Supplement 2): S699–S702. Publisher Full Text\n\nTalpur AD, Memon AJ, Khan MI, et al.: A novel of gut pathogenic bacteria of blue swimming crab Portunus pelagicus (Linneaus, 1758) and pathogenicity of Vibrio harveyi a transmission agent in larval culture under hatchery conditions. Res J Appl Sci. 2011a; 6(2): 116–127. Publisher Full Text\n\nTalpur AD, Memon AJ, Khan MI, et al.: Pathogenicity and Antibiotic Sensitivity of Pathogenic Flora Associated with the Gut of Blue Swimming Crab, Portunus pelagicus (Linnaeus, 1857). Journal of Animal and Veterinary Advances. 2011b; 10(16): 2106–2119. Publisher Full Text\n\nHamasaki K, Suprayudi MA, Takeuchi T: Mass mortality during metamorphosis to megalops in the seed production of mud crab Scylla serrata (Crustacea, Decapoda, Portunidae). Fish Sci. 2002; 68: 1226–1232. Reference Source\n\nSoundarapandian P, Thamizhazhagan E, Samuel NJ: Seed production of commercially important blue swimming crab Portunus pelagicus (Linnaeus). J Fish Aquat Sci. 2007; 2(4): 302–309. Publisher Full Text\n\nMillamena OM, Quinitio E: The effects of diets on reproductive performance of eyestalk ablated and intact mud crab Scylla serrata. Aquaculture. 2000; 181(1–2): 81–90. Publisher Full Text\n\nSouthgate P: Feeds and feed production. In: Aquaculture: farming aquatic animals and plants. Lucas J., Southgate P. (eds), Blackwell Publishing, Victoria, Australia, 2003; 172–198. Reference Source\n\nFujaya Y: Growth and molting of mud crab administered by different doses of vitomolt. Jurnal Akuakultur Indonesia. 2011; 10(1): 24–28. Publisher Full Text\n\nEfrizal, Zakaria IJ, dan Bulanin U: Pengaruh kombinasi dan level pakan alami terhadap kelangsungan hidup dan perkembangan larva Rajungan, Portunus pelagicus (Linnaeus, 1758) secara terkontrol. In. Sipahutar, H., Restuati, M., Nasution, Y., Silitonga, M., dan Gultom., E.S., Editor. Prosiding Seminar Nasional Dalam Rangka Semirata BKS-PTN Wilayah Barat Bidang MIPA Tahun 2012. Fakultas Matematika dan Ilmu Pengetahuan Alam Universitas Negeri Medan. 2012; 51–58.\n\nEfrizal: Effects of Salinity on Survival Rate and Larval Development of Blue Swimming Crab Portunus pelagicus (Linnaeus, 1758) Under Laboratory Conditions. Aust J Basic Appl Sci. 2015; 9(7): 557–563. Reference Source\n\nEfrizal A, Arshad MS, Kamarudin CR, et al.: Some Aspects of Reproductive Biology of Blue Swimming Crab Portunus pelagicus (Linnaeus, 1758) Under Laboratory Conditions. J Fish Aquat Sci. 2015; 10(2): 77–91. Publisher Full Text\n\nEfrizal: Perkembang Gonad Induk Rajungan, Portunus pelagicus (Linnaeus, 1758), dengan Manipulasi Pakan Alami dan Buatan. dalam E. Mukhtar, Syamsuardi, T., Partomohardjo dan E. Munir, Editor. Prosiding Seminar Nasional. Biodiversitas dan Ekologi Tropika Indonesia (BioETI) Universitas Andalas, Padang, 14 September 2013. 2013; 45–51. Reference Source\n\nAslamyah S, dan Fujaya Y: Effectiveness of artificial diet enriched by spinach extract on molting stimulation to produce soft shell crab. Jurnal akuakultur Indonesia. 2011; 10(1): 8–16. Publisher Full Text\n\nEfrizal: Weight-Length and Width, and Length-Width Relationships of Juvenile Blue Swimming Crab, Portunus pelagicus (Linnaeus, 1758), from Sungai Pulai Seagrass Beds, Johor, Malaysia. Aust J Basic Appl Sci. 2014; 8(7): 277–286. Reference Source\n\nAPHA: Standard Methods for examination of water and wastewater. Washington, American Public Health Association; 2012; 1360. Reference Source\n\nSteel RDG, Torrie JH: Prinsip dan prosedur statistic. PT. Gramedia Jakarta. halaman. 1990; 748. Reference Source\n\nGutiérrez-Yurrita PJ, Montes C: Bioenergetics of juveniles of red swamp crayfish (Procambarus clarkii). Comparative Biochemistry and Physiology - A Molecular and Integrative Physiology. 2001; 130(1): 29–38. Publisher Full Text\n\nGunamalai V, Kirubagaran R, Subramoniam T: Sequesration of ecdisteroid hormon into the ovary of the mole crab, Emerita asitica. Curr Sci. 2003; 85(485): 493–496. Reference Source\n\nKuballa AV, Holton TA, Paterson B, et al.: Moult cycle specific differential gene expression profiling of the crab Portunus pelagicus. BMC Genomics. 2011; 12: 147. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWahyuningsih SA: Pengaruh Dosis Penyuntikan Vitomolt terhadap Molting Kepiting Bakau (Scylla olivaceous). Skripsi Fakultas Ilmu Kelautan dan Perikanan Universitas Hasanuddin. Makassar. 2008.\n\nSusanti H: Pengaruh Dosis Vitomolt dalam Pakan Kepiting Bakau (Scylla olivacea) terhadap Molting. Skripsi. Program Studi Budidaya Perairan Fakultas Ilmu Kelautan dan Perikanan. Universitas Hasanuddin. 2009.\n\nTecha S, Chung JS: Ecdysteroids regulate the levels of Molt-Inhibiting Hormone (MIH) expression in the blue crab, Callinectes sapidus. PLoS One. 2015; 10(4): e0117278. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPreston-Mafham J, Dinan L: Phytoecdysteroid levels and distribution during development in Limnanthes alba Hartw. ex Benth. (Limnanthaceae). Z Naturforsch C. 2005; 57(1–2): 144–152. PubMed Abstract \| Publisher Full Text\n\nLafont R, Dinan L: Practical uses for ecdysteroids in mammals including humans: an update. J Insect Sci. 2003; 3: 7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHabashy MM, Hassan MMS: Effects of temperature and salinity on growth and reproduction of the freshwater prawn, Macrobrachium rosenbergii (Crustacea- Decapoda) in Egypt. International Journal of Environmental Science and Engineering (IJESE). 2011; 1: 83–90. Reference Source\n\nVijayan KK, Diwan AD: Influence of temperature, salinity, pH and light on molting and growth in the Indian white prawn Penaeus indicus (Crustacean: Decapoda: Penaeidae) under laboratory conditions. Asian Fish Sci. 1995; 8: 63–72. Reference Source\n\nJia XY, Zhuang P, Feng GP, et al.: The relationship between hemolymph bio-chemical parameters and salinity in female parent Chinese mitten crab (Eriocheir sinensis). Journal of Fisheries of China. 2012; 36(1): 91–97, (in Chinese with English abstract). Publisher Full Text\n\nLu J, Zhuang P, Feng GP, et al.: Response of osmoregulation and antioxidation system to water salinity in Parent Chinese mitten crab (Eriocheir sinensis). Mar Fish. 2011; 33(1): 39–45. (in Chinese with English abstract). Publisher Full Text\n\nCahu C, Fakhfakh M, Quazuguel P: Effect of Dietary a-tocopherol level on Reproduction of Penaeus indicus. Larvi '91: Fish & Crustasean Larviculture Symposium. European Aquaculture Society. 1991; 15: 242–244.\n\nDhawan RM, Divivedi SN, Rajamanickam GV: Ecology of the blue crab, Portunus pelagicus (Linnaeus) and its potensial fishery in Zuary Estuary. Indian Journal of Fisheries. 1976; 23: 57–64. Reference Source\n\nMeagher TD: Ecology of the crab Portunus pelagicus in South Western Australia. University of Western Australia, Australia. Ph. D. Thesis. 1971; 232. Reference Source\n\nSEAFDEC: Mudcrab culture. Asian Aquaculture. 1997; 19: 10–2.\n\nDodd LF, Grabowski JH, Piehler MF, et al.: Ocean acidification impairs crab foraging behaviour. Proc R Soc B. 2015; 282: 20150333. Publisher Full Text\n\nWayne DS, Michael AP, Glen SS: Reproduction and Growth of the Commercial Sand Crab, Portunus pelagicus (L.) in Moreton Bay, Queensland. Asian Fisheries Science. 1994; 7: 103–113. Reference Source\n\nPedapoli S, Ramudu KR: Effect of water quality parameters on growth and survivability of mud crab (Scylla tranquebarica) in grow out culture at Kakinada coast, Andhra Pradesh. Int J Fish Aquat Stud. 2014; 2(2): 163–166. Reference Source\n\nLiao IC, Murai T: Effects of dissolved oxygen, temperature and salinity on the oxygen consumption of the grass shrimp P. monodon. Maclean JL, Dizon LB, Hosillos LB, eds. The First Asian Fisheries Forum: proceedings; 1986 May 26-31; Manila, Philippines. Manila: Asian Fisheries Society, 1986; 633–636.\n\nSanares RC, Katase SA, Fast AW, et al.: Water quality dynamics in brackishwater shrimp ponds with artificial aeration and circulation. In: First Asian Fisheries Forum. (eds. J.N. Maclene, L.B. Dizon and L.V. Hosillos). Asian Fisheries Soc., Manila, Philippines: 1986; 83–86. Reference Source\n\nBoyd CE, Tucker CS: Pond aquaculture water quality management. Kluwer Academic Publisher, Massachusetts, 1998; 700. Publisher Full Text" }	[ { "id": "52710", "date": "24 Sep 2019", "name": "Rudy Agung Nugroho", "expertise": [ "Reviewer Expertise Animal physiology", "fish nutrition", "fish immunology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscripts is quite interesting but need more improvement in some parts, as shown in the annotated manuscript, and with the comments below:\n\nAbstract:\nMethods should be more details, results and conclusion must be rewritten to meet the research aims.\n\nIntroduction:\nSufficient.\n\nMethods:\nMore details about the food (Please see manuscript).\n\nThere is a possibility of residual ethanol in the sample when the ethanol used to mix amaranth and feed. How can you conclude the results with this concern?\n\nIs there any information about proximate composition of artificial feed?\n\nResults\nPlease add more information about the initial weight of crab?. Was it homogen statistically at the day 0?\n\nHow many crab did author measure for growth per replication?\n\nIt is stated in the methods that the growth results was analyzed using Anova and Duncan post hoc. However, there is no statistic information in the results. Thus it is quite difficult to determine the differences among the groups of doses.\n\nNo need to put survival rate in the table (Table 4) because all survival 100%, Table 4 should be deleted.\n\nPlease check some typos in the table. Ex: avarage change with average.\nDiscussion\nThe Discussion part can be much stronger to add why amaranth involves in decreasing or increasing growth performance\nConclusion\nIt is more likely only repeating the results. Please rewrite.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "55042", "date": "21 Oct 2019", "name": "Jirawat Saetan", "expertise": [ "Reviewer Expertise Crustacean and mollusc neuro-endocrinology", "neuroanatomy", "and reproduction" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nComment to the authors and editor:\nAbstract:-\nIntroduction:\nAuthors should give more information on the vitomolt and also any other plants that contain natural active molting compounds. Examples of testing those compounds in crustaceans should be mentioned.\nMethods:\nHow the authors differentiated crab at stage II ovary from those at other ovarian stages? Typically, the ovary is not seen from outside.\n\nAuthors could not say \"5 replications\" since the number of 5 was a number of crabs for each treatment (n = 5). The replication will be used if the authors repeat the whole treatment.\n\nHow did the authors be sure for the concentration of amaranth extract used in this study. How the spraying of 20 mL/kg of 80% EtOH solution onto the test feed of 250 ng/g treatment differed from 0, 500 and 750 ng/g treatments?\nResults:\nIf possible, the authors should start the experiment with the same crab size in each treatment. In table 1, crabs started at 100+ g BW naturally had higher growth curve than those started at 200+ g. Sensitivity to the amaranth extract between small and big crabs might not be equal! These factors might also affect the trend of result rather than the dose of amaranth extract itself.\n\nDue to the number of 5 in each treatment with some variations in your result, I am not sure that the regression with low R2 value was best in describing the relationships between the dose of extract and tested parameters.\n\nTable 4 was not necessary. Simple statement mentioning the survival rate of crabs in each treatment might be enough.\n\nTable 5 was not necessary. Almost shown parameters were mentioned already in method section!\n\nIf applicable, the data on molting cycle for all treatments should be added.\nDiscussion:\nAuthors may provide more evidences on natural ecdysteroid, vitomolt, as well as stimulating effect of these natural compound on crustacean growth and molting.\n\nBy feeding, authors should discuss how ecdysteroid, when consumed by crab, was able to be absorbed into the bloodstream.\n\nParagraph discussing the culture parameters (pH, salinity,...) was not necessary and made this manuscript weaker. Authors should discuss to sharply find out why the amaranth extract promote growth and molting and how about the same group of compound (in other plants or in nature) effects in other crustaceans.\n\nDiscussion may include the possibility of practical use for this extract in massive crab aquaculture.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-1466
https://f1000research.com/articles/8-1460/v1	19 Aug 19	{ "type": "Data Note", "title": "Curation of an intensive care research dataset from routinely collected patient data in an NHS trust.", "authors": [ "Chris McWilliams", "Joshua Inoue", "Philip Wadey", "Graeme Palmer", "Raul Santos-Rodriguez", "Christopher Bourdeaux", "Joshua Inoue", "Philip Wadey", "Graeme Palmer", "Raul Santos-Rodriguez" ], "abstract": "In this data note we provide the details of a research database of 4831 adult intensive care patients who were treated in the Bristol Royal Infirmary, UK between 2015 and 2019. The purposes of this publication are to describe the dataset for external researchers who may be interested in making use of it, and to detail the methods used to curate the dataset in order to help other intensive care units make secondary use of their routinely collected data. The curation involves linkage between two critical care datasets within our hospital and the accompanying code is available online. For reasons of data privacy the data cannot be shared without researchers obtaining appropriate ethical consents. In the future we hope to obtain a data sharing agreement in order to publicly share the de-identified data, and to link our data with other intensive care units who use a Philips clinical information system.", "keywords": [ "Intensive care", "electronic health record", "medical database", "research data", "critical care data", "ICNARC", "Philips", "clinical information system" ], "content": "Introduction\n\nThe increasing use of clinical information systems on intensive care units (ICUs) means that large amounts of patient data are being generated as part of routine care. These data are stored in electronic health records (EHR) and represent a valuable resource with huge potential to improve patient care. Collaboration between clinicians, researchers and industry stakeholders is required to realise the potential of these data by developing new methodologies and digital technologies. However, there exists a more fundamental set of barriers to making the required data available for secondary use and until these barriers are overcome the ability to maximise patient benefit via data-driven approaches will be limited. Here we introduce what we see as the four main barriers, and then explain how the publication of this data note (and its associated methodology for data curation) contributes to overcoming these barriers.\n\nThere is no standard format for storing intensive care EHR data. This is mainly due to two factors: differences between the proprietary formats used by different clinical information systems, and the high level of configurability of each system. EHR data are stored in proprietary formats designed by the companies who provide the data collection and storage software. In the intensive care units at our hospital we use the Philips ICCA clinical information system (CIS), which is currently the most widely deployed system across the NHS, with installation at 27 sites at the time of writing. Although the various available critical care CIS products do facilitate secondary data usage to some extent, they were all designed primarily as charting systems and therefore secondary use of the data is always a challenge. The main issue with ICCA is the high level of configurability of the system, meaning that data encoding can vary extensively between sites but can also change over time at a single site. The consequence of this configurability is that it can be challenging to locate and harmonise even a single simple data element, such as heart rate, for a cohort of patients over a period of time.\n\nThere are two related issues around data linkage: 1) different types of data from different sources within the hospital (or beyond) need to be linked in order to make the data more useful to researchers; and 2) data from different hospitals need to be combined to increase data volume and therefore statistical power.\n\nThe first issue relates to the scope of individual data sources. The ICCA database contains data collected routinely as part of patient care on ICU, but does not contain any information about what happened to the patient before or after their ICU stay. Therefore, taken in isolation, the data in ICCA are of limited use for research purposes. In order to make the data more useful they must be linked to other datasets that capture diagnoses, past medical history, outcomes etc. For this purpose we use data that is compiled locally for national audit by ICNARC (see Methods for more details). Linkage of our ICCA data to the local ICNARC data is a procedure that should be simple but is in fact challenging because of several error sources relating to the way that the data are collected. Developing a robust data linkage procedure has required an intimate knowledge of the data. Exposition of this data linkage procedure is one of the main purposes of this paper, because it will help other NHS trusts unlock secondary value from their data.\n\nThe second issue relates to the fact that individual intensive care datasets are relatively small. The general intensive care unit at UHB has 20 beds and treats around 1300 patients each year. To date the research database contains 4831 patients database and this number will increase to ~ 6100 with the update at the end of 2019. Most machine learning algorithms need more cases than this to achieve good performance, hence the motivation to link datasets across hospitals. Two US-based critical care datasets have achieved high volumes of data via different means. MIMIC-III1 contains around 60,000 ICU admissions, collected from a single large teaching hospital with multiple units over a period of 12 years. Conversely the eICU database2, produced by Philips, contains around 200,000 patient stays from different hospitals over a period of two yearsa. The eICU data were collected with purpose built software to facilitate high-frequency data collection in a coherent format. Both the MIMIC and eICU datasets are publicly available and their widespread use by researchers will be hugely beneficial to patients. In the UK the CCHIC3 has work on linking data from multiple hospitals with different CIS products. The challenges posed by linking data from the different proprietary systems are significant, but the data has begun to be used by researchers affiliated with the CCHIC. We feel that focusing solely on data from a single CIS system (e.g. ICCA) would significantly simplify the linkage process and that, given the widespread deployment of ICCA across the NHS, there is good potential to produce a large high-quality intensive care research database by linking data from ICCA sites only. The first stage in this process is to encourage and facilitate local pre-processing of the data at each site.\n\nThere is a growing consensus that the best way to unlock value from data is to share them widely and openly with researchers. Given the sensitive nature of medical data there are important ethical issues to consider in this context. However, we are ultimately of the opinion that it is unethical not to use routinely collected data to improve patient care. Therefore, addressing the issues around data privacy requires the development of information governance frameworks to facilitate data sharing while ensuring transparency, trust and safeguarding of patient data. The public data sharing agreements of MIMIC and the eICU represent precedents in this area that the NHS should pursue in order to unlock maximum value from their data.\n\nIn this data note we outline the steps we have taken to make our routinely collected critical care data ‘research ready’ and provide some related resources via GitHub. Our intention is that this will contribute to overcoming the above barriers, particularly by facilitating other ICUs with the ICCA system to link and process their data for secondary use. Curating our data using the methods described here has expanded our capacity for clinical reporting. We now regularly review a wide range of practices such as proning, pressure area care and prescribing. In real-time we use clinical dashboards to show the status of beds on the unit and generate retrospective reports to study trends over time. We have previously published work on the effectiveness of our clinical dashboards in improving ventilation practice via behavioural nudges4,5. Since then we have continued to expand the capabilities of the dashboards to support clinical decision making and improve the quality of care. We have collaborated with Philips on the development of dashboard intervention for acute kidney injury6,7 and have begun to explore machine learning methods for the automatic classification of ward-dischargeable patients8.\n\nIn the future, under the correct information governance framework, linkage between several ICUs with ICCA could produce a large high quality critical care research dataset. In the meantime we encourage researchers to consider using our data by obtaining the appropriate ethical consents (see Data availability) and provide a brief summary of the data that would be available to them.\n\n\nMaterials and methods\n\nIn this section we describe the processing that we have done so far to make our routinely collected data ‘research ready’. We first detail the two sources of our research data, then outline the procedure for linking data from these two sources and finally discuss the importance of further processing, including data harmonisation, to increase the general usability of these data. In the text we refer to open-source SQL and Python scripts that we have shared on our group GitHub account for readers wanting to process their own data in a similar way.\n\nICCA. Philips IntelliSpace Critical Care and Anesthesia information system (ICCA) is a patient monitoring, documentation and prescribing system used in the four intensive care units at our hospitalb. ICCA collects rich data about a patient’s condition, both via automated data streams from bedside monitors and manually input by health care providers. These data include ventilation details, medications and regular notes from medical staff. The data are stored in a reporting database, which is managed using Microsoft SQL Server and follows a star-schema that is well documented by Philips.\n\nThe ICCA data are used by medical staff to monitor patients while they are on the unit, and secondary usage has traditionally focused on financial reporting within the trust to capture the value of care provided in each ICU stay. More recently we have started to make use of the data for clinical reporting and have established regular meetings to schedule work on reporting requests from clinicians.\n\nICNARC. The Intensive Care National Audit and Research Centre (ICNARC) is an independent national charity set up with funding from the Department for Health and the Welsh Health Common Services Authority in 1993. The Case Mix Programme (CMP)9 started in 1994 is one of ICNARC’s main national audits which today provides a comprehensive dataset across 268 critical care unit, covering 99% of all adult critical care units in the in the UK and Northern Ireland. The CMP dataset (currently version 3.1) consists of 209 data fields (as listed Table S1, Extended data10), which overlap with most of the 34 data fields in the Critical Care Minimum dataset11 and include the CCMDS subset of all 14 mandatory data fields used to generate the Healthcare Resource Group (HRG). This data is collected for every patient that passes through a CMP participating ICU and covers: basic demographic information; pre-admission details including past medical history and reason for ITU admission (using the ICNARC Coding Method); severity during the first 24 hours; number of days of organ support during their ICU stay and outcomes on both leaving the unit and then final discharge from hospital. The purpose of the audit is to provide a national resource for research and a local and national benchmarking tool for individual critical care units.\n\nWard Watcher12 is the bespoke proprietary software (provided by Critical Care Audit Ltd) we use in the trust to collect this CMP dataset before sending it off to ICNARC. This software allows us to collect extra information for each patient that is not sent to ICNARC but is used within the Trust to generate detailed custom reports. It has been configured to automatically generate new records when a new admission is entered into a bed space on the Philips ICCA system and will pull data from the flowsheet and completed forms in ICCA for manual verification.\n\nA careful procedure is required to link datasets from different sources to produce valid and usable data. Here we describe our procedure for linking data from ICCA and ICNARC to produce patient records with both routinely collected ICU data and outcome descriptors. This method will be useful for any intensive care unit the ICCA system who want to make secondary use of their data in-house. The method is also detailed step-by-step in an iPython notebook (see Script S1, Software availability10).\n\nThe main challenge to overcome is that erroneous entries in both datasets prevent a clean link. Without these errors the linkage would be a simple case of joining data tables on a unique identifier corresponding to each ICU stay. Therefore, we must first identify the erroneous entries and handle them according to the type of error that produced them. This procedure would not be possible without an intimate first-hand knowledge of the data and they way they are generated. There are three stages in the data linkage: first we handle the errors in the ICNARC data, then we handle the errors in the ICCA data and finally we link the two datasets together.\n\nEvery patient record in the ICNARC datac is manually validated by the data team, so we can be sure that each record corresponds to a real ICU stay and contains valid patient data. In the Ward Watcher software each ICNARC patient record links to an identifier in ICCA called the encounterId. In theory the encounterId uniquely identifies each ICU stay that has been captured in the CIS. However, there are various sources of error in the ICCA encounterIds which break the one-to-one mapping with patient records in Ward Watcher. For a small number of cases the patient record in Ward Watcher points to an empty or corrupt ICU stay in ICCA. In these cases we simply redirect the record in Ward Watcher to point to the correct stay in ICCA. For completeness we also create a new column to record the erroneous ICU stay that was pointed to originally.\n\nWhen patients are admitted to ICU, a record with a unique encounterId is manually created in ICCA. All data associated with that ICU stay is linked with this encounterId until the patient is discharged from ICU, at which point they are manually removed from the system. Since the admission and discharge actions in ICCA are conducted manually and are not retrospectively validated, there is potential for a number of different types of error. For example, patients can be admitted and discharged erroneously leading to phantom, nested or disjointed stays. All the potential types of error are listed in Table S2 (Extended data10), but there are broadly two classes of error, which are handled differently: 1) multiple encounterIds corresponding to a single ICU stay; and 2) multiple actual ICU stays with a single encounterId. For the first class of error, we replace the duplicate encounterIds with the original encounterId that was created for that stay such that a single coherent record is produced. We again produce a new column (specifically in the D_Encounters table) to record the duplicate encounterIds that have been replaced. For the second class of error there is no simple solution that could be robustly automated, so we leave these cases for manual processing by individual researchersd. To facilitate manual processing we introduce another column (to the table D_Encounters) which specifies the type of error, if any, associated with each encounterId.\n\nHaving handled the errors in both datasets, we now have one-to-one mapping between ICNARC records and stays in ICCA. We then extract all the CMP patient data from Ward Watcher in a standard XML format and use it to produce another table in our research database called D_Icnarc. This table has one row for each ICU stay and one column for each of the 209 variables in the CMP dataset, and links to other tables via encounterId and ptCensusIde.\n\nThe configurability of ICCA means that the way interventions are encoded can change over time. For retrospective studies it is necessary to search for medical concepts and variables in the SQL database, which can be time consuming. We have provided a well commented SQL script (see Script S2, Software availability10) for locating variables in the back end of ICCA which should be useful for anyone working with the system. In general the best strategy is to search on the longLabel for interventions and on the shortLabel for the corresponding attributes, and then to calculate usage frequency to confirm that the variable located is in use. In the future we hope to produce a software tool for variable location that is usable by those without knowledge of SQL or experience of working with ICCA.\n\nThe full database is stored on a secure hospital server to which only UHB data managers have access. We follow the guidelines of the NHS Health Research Agency Confidentiality Advisory Group13. Curation of the data for internal audit and service evaluation does not require research ethics approval, and for projects that extend beyond routine reporting we produce de-identified extracts of the required data with sensitive information removed (names, dates of birth, addresses, rare diagnoses, etc.).\n\n\nDataset validation\n\nThe ICNARC data are validated internally at our hospital and externally at the national office. Therefore, we can have confidence in the validity of these data. The above procedure for data linkage also removes erroneous entries in the ICCA data. Users of the data must be aware that there are other sources of error in CIS data. In particular, some data are entered manually (medical notes, free form laboratory results, etc.) and are therefore vulnerable to corruption. Certain data fields are populated automatically (e.g. from bedside monitors) but not stored until a nurse confirms that the value is representative. Such fields are therefore valid when recorded but subject to missing values.\n\nIn Table 1 we provide a brief summary of 30 selected physiological variables to give readers a feel for the type of data contained in the database, including the frequency of recording of different variables and the extent of missing data values. We also provide a demographic summary of the patients represented in the data (Table 2). Readers are referred to Supplementary Figures S1–S4 (see Extended data10) for further demographic information, and Supplementary Figures S5–S7 for distributions of variable values.\n\n‘Record completeness’ is the percentage of ICU stays that contain at least one recording of the variable. ‘Frequency recorded’ is the number of times the variable is recorded per hour for the ICU stays that contain records of that variable. (Note: these frequencies are calculated over the full length of stay and so may be distorted when a variable is measured only during a subset of the stay.)\n\n\nFuture work\n\nThe curation of this data has highlighted to us the importance of close collaboration between the people and teams responsible for collecting, administering and validating the data. The more that is known about an intensive care dataset—the way the data are collected, the way they are affected by clinical practice, idiosyncrasies in the digital systems involved, operational factors—the more value and information that can be extracted from them and ultimately the more value we can deliver to patients. In the future we will continue to improve and expand this research database. In particular we will work with colleagues in NICU, PICU and CICU to link data from the other intensive care units in our hospital. We will also look to include datasets from across the trust to capture information about patient hospital admissions outside the ICU.\n\nWe hope to work with external collaborators to develop a robust method for de-identifying medical notes. Finally, we will explore the possibility of linking with data from external NHS trusts who also use ICCA in their ICUs. Eventually the expansion of this research data will require more extensive data harmonisation to combine multiply-defined clinical concepts, and crucially will require a bespoke information governance framework to allow us to bring this data to researchers. We note that there is a precedent for such governance agreements in other projects referenced previously1–3.\n\n\nData availability\n\nThe sensitive nature of these data means that they are only available internally to UHB staff for the purposes of clinical audit and service evaluation activities via the CAG guidelines. For external researchers, ethical approval may be obtained via formal application to the NHS Integrated Research Application System (IRAS) for a specific research project. The IRAS website (www.myresearchproject.org.uk) has full instructions; however, interested parties are advised to contact the corresponding author (christopher.bourdeaux@uhbristol.nhs.uk) to discuss the application.\n\nZenodo: UHBristolDataScience/data-note-extended-data https://doi.org/10.5281/zenodo.336128714.\n\nThis project contains the following extended data:\n\nTable S1: extended_tables/icnarc_cmp_dataset_properties.xlsx\n\nTable S2: extended_tables/icca_encounterid_error_types.xlsx\n\nFigure S1: extended_figures/admisson_types_discharge_reasons.png\n\nFigure S2: extended_figures/discharge_time_histograms.png\n\nFigure S3: extended_figures/reasons_for_admission.png\n\nFigure S4: extended_figures/stay_length_histograms.png\n\nFigures S5–S7: extended_figures/variable_hists[1-3].png\n\nExtended data are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).\n\n\nSoftware availability\n\nScript S1 available from: GitHub (file ‘clean_encounterids.py’).\n\nScript S2 available from: GitHub (file ‘variable_location_in_ICCA.sql’).\n\nArchived code at time of publication: https://doi.org/10.5281/zenodo.335875010.\n\nLicence: MIT licence.\n\n\nNotes\n\naThis is the publicly available component of the eICU dataset. The full dataset held by Philips is much larger.\n\nbThe use of the same database by the four units is one source of error in the data (e.g. erroneous transfers or patients being attached to the wrong unit identifier).\n\ncNote that in some very rare cases there are stays which are excluded from the ICNARC data.\n\ndFor example, researchers may wish to simply remove such cases, although removal would likely introduce some bias since these cases usually represent readmissions to ICU. Alternatively they may wish to manually split the stay into two records.\n\neThe ptCensusId in ICCA uniquely identifies spells in different units during the same ICU stay.", "appendix": "Author contributions\n\n\n\nGP, PW, JI and CM together extracted and pre-processed the data. GP, JI, PW and CB provided intimate knowledge of data collection procedures and systems. RS and CB conceived of the research dataset and oversaw its curation. CM conducted the coding and analysis. All authors contributed to writing the manuscript and approved the final version.\n\n\nGrant information\n\nCM was funded by the EPSRC Impact Acceleration Account (EP/R511663/1) with a contribution from Above and Beyond.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe would like to thank all of our colleagues within UHB who have made these data curation activities possible and who continue to support secondary use of the data. In particular: Russell McDonald-Bell, Matt Rogers, Colin Salandy, Amy Smith, all the members of UHBDataScience and our colleagues in NICU, PICU and CICU who work with ICCA. The expertise of Brian Millar (author of the Ward Watcher software) and Phil Stuart-Douek (Philips) has also been essential.\n\n\nReferences\n\nJohnson AE, Pollard TJ, Shen L, et al.: MIMIC-III, a freely accessible critical care database. Sci Data. 2016; 3: 160035. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPollard TJ, Johnson AEW, Raffa JD, et al.: The eICU Collaborative Research Database, a freely available multi-center database for critical care research. Sci Data. Nature Publishing Group, 2018; 5: 180178. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHarris S, Shi S, Brealey D, et al.: Critical Care Health Informatics Collaborative (CCHIC): Data, tools and methods for reproducible research: A multi-centre UK intensive care database. Int J Med Inform. 2018; 112: 82–89. PubMed Abstract \| Publisher Full Text\n\nBourdeaux CP, Thomas MJ, Gould TH, et al.: Increasing compliance with low tidal volume ventilation in the ICU with two nudge-based interventions: evaluation through intervention time-series analyses. BMJ Open. 2016; 6(5): e010129. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBourdeaux CP, Birnie K, Trickey A, et al.: Evaluation of an intervention to reduce tidal volumes in ventilated ICU patients. Br J Anaesth. 2015; 115(2): 244–251. PubMed Abstract \| Publisher Full Text\n\nPachucki M, Ghosh E, Palanisamy K, et al.: 1354: Acute kidney injury (aki) progression during the first five days of an icu stay. Critical Care Medicine. 2018; 46(1): 660. Publisher Full Text\n\nPachucki MA, Ghosh E, Eshelman L, et al.: Descriptive study of differences in acute kidney injury progression patterns in general and cardiac intensive care units. Journal of the Intensive Care Society. 2018; 1751143718771261. Publisher Full Text\n\nMcWilliams CJ, Lawson DJ, Santos-Rodriguez R, et al.: Towards a decision support tool for intensive care discharge: machine learning algorithm development using electronic healthcare data from MIMIC-III and Bristol, UK. BMJ Open. 2019; 9(3): e025925. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nICNARC: A brief history of the cmp. URL 2019. Reference Source\n\nMcWilliams C: UHBristolDataScience/ICNARC-to- Philips-Linkage: Software resources for data curation. 2019. http://www.doi.org/10.5281/zenodo.3358750\n\nNHS Digital: Critical care minimum data set overview. 2019. Reference Source\n\nMillar B: Ward watcher. 2014. Reference Source\n\nHealth Research Authority: Guidance for cag applicants. 2019. Reference Source\n\nMcWilliams C, Inoue J: UHBristolDataScience/data-note-extended-data. 2019. http://www.doi.org/10.5281/zenodo.3361287" }	[ { "id": "52676", "date": "27 Aug 2019", "name": "J Duncan Young", "expertise": [ "Reviewer Expertise My expertise is in critical care and associated interrogation of routinely collected healthcare data. I have no expertise in coding and have made this clear in the report. Please note the wording below (“I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard”) has been added to my peer review report by the publishers and was neither a part of the report I submitted nor do I have any control over this addition." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nGeneral comments\nThank you for the opportunity to review this manuscript. I have reviewed the text and figures. However I am not an expert on IT/computing and so have not commented on the details of the SQL and PYTHON code in the repositories.\nThis paper describes the curation and linking of two databases containing information on patients treated on intensive care units in Bristol. The first database (ICCA) contains detailed information on patients’ stay in a single Trust’s intensive care unit collected as part of routine care. The second contains data submitted on the same patients to a national comparative audit programme (ICNARC CMP). The manuscript is well written and is clear.\nSimilar challenges have been addressed elsewhere (notably by the CCHIC teams in the UK and the MIMIC-III team in the USA), though there is very little detailed information on the processes and problems these teams encountered which has been published. This paper addresses some of the lack of detail.\nThe paper contains descriptions of the curation and linking processes. There are no data linked directly to the paper, as publishing identifiable patient data is not permitted. However, contact details are given to allow interested researchers to explore obtaining appropriate permissions to interrogate the data.\nIn general the paper is informative and useful. It might benefit from a brief comment on how generalisable these methods are to other patient groups where highly granular data are collected such as patients treated in Emergency departments, or those undergoing surgery or invasive procedures.\n\nMinor detailed comments\nThe sentence on ICNARC’s origin should probably be modified to read “The Intensive Care National Audit and Research Centre (ICNARC) is an independent national charity originally set up with funding from the Department for Health and the Welsh Health Common Services Authority in 1993” as funding now comes from different sources.\n“Barrier 3: Data privacy”. The MIMIC-III and eICU programmes are able to share data publically and they are anonymised. There is no mention of this approach and the difficulties with true anonymisation, this paper rather assumes data will be accessed using ethical approvals.\nIt might be helpful to emphasise that the XML file format that Wardwatcher software uses to export ICNARC CMP data is common to all the different software packages used to collect ICNARC data, and is not a software-specific format.\nUse of intensive care as adjective (eg “…intensive care EHR data”) is common in published papers but is probably best avoided.\nSupplementary material graphics comments\nDischarge reasons bar chart: No X axis labels.\n\nDischarge time histogram: X axis labels at 5h 33m 20s intervals. Why this unusual spacing?\n\nStay length: Unusual to use logged Y axis for these graphs though I assume this is because of the high frequency of single day stays.\n\nVariables histogram 1: FiO2 is fractional, not %. Units needed for heart rate, haemoglobin, respiratory rate, SpO2, and blood pressures on X axes.\n\nVariable histograms 2&3: Attention to all X axis units as above. SI notation for partial pressures (P02, PCO2) uses a capitalised “P”.\n\nIs the rationale for creating the dataset(s) clearly described? Yes\n\nAre the protocols appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and materials provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? No", "responses": [ { "c_id": "4861", "date": "29 Aug 2019", "name": "Chris McWilliams", "role": "Author Response", "response": "Professor Young - Thank you for taking the time to read and review our manuscript. We appreciate your comments and will act on your suggestions to produce a revised version once we have received the other peer review reports. We will then also provide a detailed response to your review." } ] }, { "id": "52675", "date": "03 Sep 2019", "name": "Min Ji Lee", "expertise": [ "Reviewer Expertise Intensive Care Medicine." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors describe a curated dataset of 4831 adult intensive care patients treated at the Bristol Royal Infirmary between 2015 and 2019. Two critical care data sources (ICCA and ICNARC) are linked and curated to create a single comprehensive ‘research ready’ dataset. By publishing the curation process the aim is to help external researchers make secondary use of their own routinely collected data. Fundamental barriers to making the required data available for secondary research use are discussed. Due to privacy constraints, the dataset is not fully published but external researchers may gain access through a formal application process.\n\nFrom the perspective of a novice data scientist and clinician this was an insightful and informative paper. The data note explains the rationale, barriers and methodologies allowing transparency and reproducibility for interested external researchers. Scripts outlining the dataset curation process were easy to follow with step-by-step commentary. Making this information accessible provides the opportunity for deeper understanding, in particular to those new to data science but curious about its potential. This is important given the need for close collaboration between clinicians, researchers and industry stakeholders to realise the full potential of routinely collected data to improve patient care.\nThe authors discuss how the publication of the data note and curation methodology contributes to overcoming the barriers of data format and data linkage. However its role in mitigating barriers associated with data privacy is less clear. Further explanation may be of interest as the tension between maintaining data privacy and usability of data for researchers is highly relevant in this field.\n\nTo illustrate the barriers related to data format, the authors describe the challenges in locating and harmonising a single data element such as heart rate within the Philips ICCA clinical information system (CIS). The high level of configurability, where data elements can be renamed and relabelled between sites, can prevent cross-site collaboration and sharing of these modifications using code review despite using the same CIS. As these factors are beyond the researcher’s direct control, we would welcome the authors’ perspective on how commercial companies could make this process easier.\n\nIs the rationale for creating the dataset(s) clearly described? Yes\n\nAre the protocols appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and materials provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? No", "responses": [] } ]	1	https://f1000research.com/articles/8-1460
https://f1000research.com/articles/7-1371/v1	31 Aug 18	{ "type": "Clinical Practice Article", "title": "Visceral leishmaniasis (Kala-azar) in Qom Province, Iran: Report of two cases", "authors": [ "Leyli Zanjirani Farahani", "Abedin Saghafipour", "Mehdi Mohebali", "Behnaz Akhoundi", "Hedayatollah Raufi", "Leyli Zanjirani Farahani", "Behnaz Akhoundi", "Hedayatollah Raufi" ], "abstract": "Visceral leishmaniasis (VL) is a fatal parasitic zoonotic worldwide disease, which transmits to humans by the infected Phlebotomine sand fly bite. The common form of VL in Iran is the Mediterranean type with the causative agent of Leishmania infantum, whose main reservoirs are stray and domesticated dogs. The disease has several endemic foci in Iran, mostly seen among children under the age of 10, living in rural areas and nomadic tribes. The first cases of Kala-Azar in Qom province, central Iran, were reported in the year 2001, from the villages of Ghahan district. After conducting VL control strategies in the area, no new cases of the disease had been reported until recently. The cases described here are two 2-year-old girls, living in the urban parts of Qom province, one of whom did not have a history of traveling to known endemic areas of the disease. The patients were admitted to hospital in 2016-2017, complaining from recurrent fever with unrecognized reason, associated with decreased appetite and weight loss. Disease follow-up demonstrated anemia and splenomegaly, which led to diagnosis of VL, and both patients are now fully recovered. VL was presumed to be controlled in Qom province but the present cases indicate that possible VL existence remains in the region. Therefore, urgent studies and periodic monitoring are needed to identify potential reservoirs of VL in the area.", "keywords": [ "Visceral leishmaniasis", "Kala-azar", "Qom" ], "content": "Introduction\n\nProtozoan parasites are important, rare, uncontrolled, human lethal pathogens that exist due to insufficient epidemiological studies and lack of adequate information on the aspects of diagnosis and treatment. Problems caused by parasitic diseases in different regions have their own special characteristics. Visceral leishmaniasis (VL) or Kala-azar is one of the most severe zoonotic diseases caused by different species of Leishmania, which leads to death in 95% of cases if not treated1. An estimated 50,000–90,000 new cases of VL occur worldwide annually2. VL has been linked to hygiene and environmental health, along with malnutrition, weak immune system and population displacement considerations. VL is commonly observed in children in the under 10 year age group, especially 1–5 year-olds, but also afflicts adults suffering from immunodeficiency3. The zoonotic form of VL is caused by Leishmania infantum whose pathogenic form transmits from the animal reservoir to humans through infected Phlebotominae sand fly bites. The parasite’s natural reservoir is dogs, wild canids, foxes, jackals and occasionally wolves4.\n\nVL is seen most commonly in rural areas and clinical symptoms vary from asymptomatic forms and restricted infection to lethal VL. Disease incubation period lasts from a few weeks to several months5. In Iran, fever and anemia have been reported as the most common clinical signs and hepatosplenomegaly is generally displayed six months after the onset of the infection. Bone marrow involvement causes severe anemia and cachexia in the patient. Finally, secondary bacterial infections can result in the patient's death. VL clinical diagnosis is difficult due to nonspecific symptoms similar to other diseases, such as malaria, typhoid fever, brucellosis, lymphoma and leukemia, especially in non-endemic regions5,6.\n\nBetween 1998 and 2012 in Iran, 2632 cases of VL were recorded, with the majority of cases in the northern and southern parts of the country. The highest number of cases were in April and July in the age group 1–3 years and the annual average over the 14-year period was 175.4 cases. While the peak incidence was recorded in 2000 (13.15% of total Leishmania cases), VL occurrence decreased in the following years. The first cases of VL was reported in Qom province villages in 2001 and no new case has been reported until recently5.\n\nIn this Clinical Practice Article, two cases of Kala-azar are reported, which were detected in Hazrat-E-Masoumeh Hospital in Qom Province, Iran, in 2016 and 2017.\n\n\nCase 1\n\nIn February 2016 a 22-month-old girl, who was living in Qom’s downtown, was admitted to Hazrate-E-Masoume Hospital with irregular prolonged fever, cough and loss of appetite for about one month. In the initial follow-up, the cause of fever remained unrecognized and the patient was referred to the hospital, accordingly. Based on her parent’s statement, the child had travelled to Dastgerd village, in Kahak district, south of Qom Province, in November 2016, two months before clinical signs appear.\n\nIn early clinical examinations, the patient’s throat, ears, heart and lungs were functioning normally. Ultrasonography report showed normal liver tissue and splenomegaly with 14.5 mm spleen span (Figure 1). Blood smear examination showed hypochromic microcytic anemia with white blood cell and platelet number reduction (Table 1).\n\n(A) Case 1; (B) Case 2.\n\nDue to associated fever with enlargement of the spleen and pancytopenia, Direct Agglutination Test (DAT) for leishmaniasis was performed along with other serological tests, which showed positive result with a high titer of anti-L. infantum antibodies as 1:6400. To confirm, Immunofluorescence Assay (IFA) demonstrated a positive result too (IFA>1:640).\n\nAfter diagnosis, Amphotericin-B injection was prescribed at 1 mg/kg for the first day, increased to 5 mg/kg during one week. The last dose was continued until day 10. As soon as treatment began, the patient’s fever reduced and the patient’s general state improved. In the next follow-up, two weeks later, the blood cell count had risen and the patient was considered successfully treated.\n\n\nCase 2\n\nIn April 2017 a 26-month-old girl was admitted to Hazrat-E-Masoumeh Hospital in Qom. The patient lived in Qom city, and had no history of travelling to VL endemic regions since she was born. The patient presented with an unknown, persistent fever, anorexia, and general weakness, which had started four months ago. The patient had some bruises on her abdomen and legs that appeared a month earlier, which caused the physicians to suspect anemia and leukemia.\n\nUltrasonography demonstrated mild enlargement of the spleen (Figure 1). Examinations showed reduction in all blood elements (Table 1). The results of typical serological tests were negative. Bone marrow aspiration was evaluated because of pancytopenia in which no blast cell was seen. Then observation of amastigotes of Leishmania parasite (Leishman-Donovan bodies) within bone marrow macrophages, and the positive DAT result (>1:3200) was observed, and visceral leishmaniasis was diagnosed. Therefore, Amphotericin-B treatment was initiated with dose of 1 mg/kg for 21 days. After four days, the patient's fever disappeared, her general condition improved and blood cell number increased at the next month follow-up.\n\n\nDiscussion\n\nVisceral leishmaniasis (VL) is endemic in some provinces of Iran, including Ardebil, East Azarbaijan, Bushehr, Fars, and North Khorasan, and sporadically occurs in other provinces. The first human VL case in Iran was observed in Northern Iran in 19493. In Qom province, the first VL cases were observed in 2001, during the study by Fakhar et al., in which 1.7% of 416 serum samples were diagnosed seropositive, and was related to 25% contamination of dogs in rural districts of Ghahan, a northwestern part of the province7.\n\nThis article reports the first cases of Kala-azar in urban areas of Qom province. So far, over 95% of the cases in Iran have been from rural areas3. Despite the history of traveling to a village in the first case presented here, the disease may not be linked to the village definitely. The probability of there being more patients in Qom province and referring them to hospitals out of the province requires further investigation. Due to the lack of knowledge about VL, the cases were referred to hospital more than two months after the onset of symptoms. Therefore, there is not enough available history of children like the presence of malnutrition or other underlying illness which may affect the immune system of cases.\n\nThe highest incidence of Kala-azar in endemic areas of Iran is reported in children ≤5 year-olds3. The average age of the patients reported in this paper was also two years-old, which is consistent with these findings. The symptoms of VL are related to involvement of the reticuloendothelial system, which includes enlargement of the spleen, liver and lymph nodes. Moreover, bone marrow infection leads to a decrease in its normal activity, resulting in anemia, leukopenia, and thrombocytopenia6. The common symptoms in the present patients were intermittent fever, pancytopenia and splenomegaly. In previous studies, the main hallmarks of the disease were persistent fever, pallor, and spleen and liver enlargement3. In the present cases no hepatomegaly was observed. Normochromic normocytic anemia and pancytopenia (decrease of all blood cell types) were features of identified anemia in patients in this report. In one quarter of VL cases, VL develops actively and symptoms appear between 2–8 months after parasite entrance to the body6,8.\n\nDAT and IFA serology methods are considered the most efficient diagnostic tests for Kala-azar. The definitive parasite diagnosis is done through parasitological methods involving microscopic examination of spleen, bone marrow and lymph nodes aspiration to observe the parasite, which are sensitive but invasive and potentially hazardous9. Evaluation of indirect IFA and DAT showed that IFA is more reliable in VL diagnosis than DAT but needs advanced laboratory equipment, while DAT is a simple, precise, cost effective and applicable test for all situations10. Due to the possibility of cross-reaction between VL and cutaneous leishmaniasis, tuberculosis and malaria in performing DAT, the history of such diseases in the patient should be investigated as well as any blood transfusion history. Upon referral of the first patient reported here, both DAT and IFA results were positive. A DAT positive result is trustworthy accompanied by clinical symptoms, but it is advised to be confirmed by using a definitive parasitology method such as bone marrow aspiration, which was performed in the second patient.\n\nIn Iran, the highest disease incidence had been recorded in autumn and winter, which is consistent with the onset of the disease in the cases reported here11. The first patient described here had travelled to a village in Qom province, but the second case had been living in Qom city since she was born. So having no history of travel to a known endemic areas of VL does not rule out the risk of the disease. In a VL reservoir study in Iran, not only stray dogs, but also domestic dogs have been infected with Leishmania parasite. In 2012–2014, asymptomatic and symptomatic domestic dogs were compared in Meshkin-shahr; 18.6% of asymptomatic domestic dogs had VL infection, and surprisingly, 13.4% of the asymptomatic dogs demonstrated negative serology tests while they had been positive in the parasitology exam. Therefore, the presence of domesticated and stray dogs in urban and rural areas plays an important role in the occurrence of sporadic cases of VL12,13. Because of the complex nature of VL manifestations and the risk of misdiagnosis, physicians in urban and rural health centers should remain vigilant, as they play a crucial role in early diagnosis and timely treatment of patients14.\n\nIn Qom Province, central Iran, Kala-azar had presumably been controlled after control strategies were implemented, and no new cases had been reported recently. The two cases in this report indicate that VL might not be eradicated totally in this area. Therefore, further studies on epidemiological aspects of Leishmania reservoirs and vectors are recommended along with increased surveillance system awareness.\n\n\nConsent\n\nWritten informed consent for the publication of clinical details of the patients described here was obtained from their parents.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.", "appendix": "Grant information\n\nThis study was part of MSc thesis No. 240/221, financially supported by Tehran University of Medical Sciences.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe thank the kindly cooperation of the officials in Hazrat-E-Masoumeh Hospital in Qom, Iran, and Qom province Health Center staff as well as Leishmaniasis Laboratory coworkers at the Medical Parasitology and Mycology Department of Tehran University of Medical Sciences, Tehran, Iran, in the preparation of this report.\n\n\nReferences\n\nWorld Health Organization: Control of the leishmaniasis: report of a meeting of the WHO Expert Committee on the Control of Leishmaniases, Geneva, March 22–26, 2010. World Health Organ Tech Rep Ser. 2010; 949: 1–186. Reference Source\n\nWorld Health Organization: Leishmaniasis page: Home/ News/ Fact sheets. 2018. Reference Source\n\nMohebali M, Edrissian GH, Shirzadi MR, et al.: An observational study on the current distribution of visceral leishmaniasis in different geographical zones of Iran and implication to health policy. Travel Med Infect Dis. 2011; 9(2): 67–74. PubMed Abstract \| Publisher Full Text\n\nYaghoobi-Ershadi M: Phlebotomine Sand Flies (Diptera: Psychodidae) in Iran and their Role on Leishmania Transmission. J Arthropod Borne Dis. 2012; 6(1): 1–17. PubMed Abstract \| Free Full Text\n\nShirzadi MR, Esfahania SB, Mohebalia M, et al.: Epidemiological status of leishmaniasis in the Islamic Republic of Iran, 1983-2012. East Mediterr Health J. 2015; 21(10): 736–42. PubMed Abstract \| Publisher Full Text\n\nMcGwire BS, Satoskar AR: Leishmaniasis: clinical syndromes and treatment. QJM. 2014; 107(1): 7–14. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFakhar M, Mohebali M, Barani M: Identification of endemic focus of kala-azar and seroepidemiological study of visceral leishmaniasis infection in human and canine in Qom province, Iran. Armaghane Danesh. 2004; 9(33): 43–52 (Persian). Reference Source\n\nChappuis F, Sundar S, Hailu A, et al.: Visceral leishmaniasis: what are the needs for diagnosis, treatment and control? Nat Rev Microbiol. 2007; 5(11): 873–82. PubMed Abstract \| Publisher Full Text\n\nMohebali M, Edrissian GhH, Nadim A, et al.: Application of direct agglutination test (DAT) for the diagnosis and seroepidemiological studies of visceral leis-hmaniasis in Iran. Iranian J Parasitol. 2006; 1(1): 15–25. Reference Source\n\nMikaeili F, Fakhar M, Sarkari B, et al.: Comparison of serological methods (ELISA, DAT and IFA) for diagnosis of visceral leishmaniasis utilizing an endemic strain. Iran J Immunol. 2007; 4(2): 116–121. PubMed Abstract\n\nAsgari Q, Fakhar M, Motazedian H: Nomadic kala-azar in south of Iran. Iranian J Publ Health. 2006; 35(3); 85–86. Reference Source\n\nShokri A, Fakhar M, Teshnizi SH: Canine visceral leishmaniasis in Iran: A systematic review and meta-analysis. Acta Trop. 2017; 165: 76–89. PubMed Abstract \| Publisher Full Text\n\nMolaei S, Dalimi A, Mohebali M, et al.: Study of Canine Visceral Leishmaniasis in Symptomatic and Asymptomatic Domestic Dogs in Meshkinshahr City, Iran. J Ardabil Univ Med Sci. 2016; 16(1): 105–115 (Persian). Reference Source\n\nMohebali M, Edrissian GH, Shirzadi MR, et al.: Integrated visceral leishmaniasis surveillance system in primary care for children in Meshkin-Shahr district, north-western Islamic Republic of Iran. East Mediterr Health J. apps.who.int. 2010; 16(10): 1050–4. PubMed Abstract \| Publisher Full Text" }	[ { "id": "37818", "date": "05 Sep 2018", "name": "Aref Salehzadeh", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe table is not acceptable in present form and should be corrected accurately by an expert physician, for example, the platelet number should be shown per ml (mm3), the leukocytes count should be present as number per ml, or hemoglobin level by g/dl.\n\nThe authors should include a figure of bone marrow smear. The bone marrow aspirate is the gold standard for confirming diagnosis. A definite diagnosis requires the demonstration of the parasites in tissue smears of the affected organs.\nThe details of patients are very brief and has been not described precisely for instance in second paragraph of discussion authors have argued that “the cases were referred to hospital more than two months after the onset of symptoms” whereas such information is not seen in description of case one in page 3. Did the authors check the possibility of cutaneous leishmaniasis, tuberculosis and malaria infections in patients?\n\nWhat about congenital transmission? Did the authors check mothers? Because there are evidences that vertical transmission might occur either transplacentally during pregnancy (in utero) or, most likely, during labor via blood exchange from the mother to the child (although in the vast majority of congenitally infected children, the symptoms and signs of VL develops during the first 12 months of life) Discussion is weak and should be rewritten and I cannot approve the manuscript in this form.\n\nIs the background of the cases’ history and progression described in sufficient detail? Partly\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Partly\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the conclusion balanced and justified on the basis of the findings? No", "responses": [ { "c_id": "4378", "date": "22 May 2019", "name": "Leyli Zanjirani Farahani", "role": "Author Response", "response": "Referee response: 1-The table is not acceptable in present form and should be corrected accurately by an expert physician, for example, the platelet number should be shown per ml (mm3), the leukocytes count should be present as number per ml, or hemoglobin level by g/dl. Many thanks for your comment, we will correct the table as you mentioned. 2-The authors should include a figure of bone marrow smear. The bone marrow aspirate is the gold standard for confirming diagnosis. A definite diagnosis requires the demonstration of the parasites in tissue smears of the affected organs. Thanks for your comment, we agree you in this matter. But because of the onset of this manuscript after the patient recovery and they were discharged from the hospital, no figure of bone marrow slide was obtained while diagnosing leishmaniasis. Since the blood smears are saved in the laboratory archive, we can photograph them and add it to the article with the editor approve. 3-The details of patients are very brief and has been not described precisely for instance in second paragraph of discussion authors have argued that “the cases were referred to hospital more than two months after the onset of symptoms” whereas such information is not seen in description of case one in page 3. Regarding your comment, the parents of patients had visited various physicians at the onset of the disease symptoms and unfortunately, they had no complete report of these visits. For this reason, the data before being admitted to the Hospital is not complete. 4-Did the authors check the possibility of cutaneous leishmaniasis, tuberculosis and malaria infections in patients? Many thanks for your precision, we have discussed this issue in paragraph 4 of the discussion. Because of the possible interaction between some diseases such as malaria, tuberculosis and cutaneous leishmaniasis and VL diagnosis via DAT, the physicians had assessed it in primary examinations. Also there was no history of Kala-azar in the patients family members. 5-What about congenital transmission? Did the authors check mothers? Because there are evidences that vertical transmission might occur either trans placental during pregnancy (in utero) or, most likely, during labor via blood exchange from the mother to the child (although in the vast majority of congenitally infected children, the symptoms and signs of VL develops during the first 12 months of life) As you mentioned, it would be important in the first 12 months of the life. But patients described here are older than two years. However, we would like to inform you that there was no history of Kala-azar in the family members of children. Respecting your comment, it should be noted that the antigen used in DAT method is prepared using Leishmania infantum, which is the agent of the Mediterranean Leishmaniasis and causes the disease in children. The probability of Mediterranean Leishmaniasis in an adult and congenital transmission with this high antibody titer at 2 years of age seems unlikely. 6-Discussion is weak and should be rewritten and I cannot approve the manuscript in this form. Thanks for your valuable guidance. We would appreciate if you could help us with the shortcomings in the article's discussion. Since this manuscript is important due to reporting VL cases in urban areas of Qom province, central Iran, which had not been reported for long, we will do our best to complete the discussion and convince you." }, { "c_id": "4717", "date": "19 Aug 2019", "name": "Leyli Zanjirani Farahani", "role": "Author Response", "response": "Dear Dr. Aref Salehzadeh, Many thanks for your precious comments on our article. Here are our respond to your latest comments: 1-Please check and correct the sentence “The role of domestic or stray dogs and Phlebotominae sand flies in the transmission cycle of the disease in clear” in line 21-22 of the fourth paragraph of the discussion section. The mistake is corrected. 2-Please check and correct the word “congenital” in line 14 of the 5th paragraph of the discussion section. We checked the word “congenital” in the dictionary and on the internet but it seems correct, respecting your opinion. 3-Please clarify which types of Amphotericin-B has been used (liposomal or …). Since the liposomal form of Amphotericin-B is known as “AmBisome®” we didn’t explain about the type of used Amphotericin-B in treatment of the patients. 4-In line 4 of the 3rd paragraph of discussion please add one or two related references. Respecting your comment, there is only a report of both patients history in the mentioned lines which does not need any reference." } ] }, { "id": "45651", "date": "25 Apr 2019", "name": "Joziana Muniz de Paiva Barçante", "expertise": [ "Reviewer Expertise Parasitology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nLeishmaniasis is considered an important zoonotic disease worldwide. So, the publication is relevant. Some aspects might be considered before the indexing of the manuscript:\n\nIs necessary a revision regarding correct biological concepts and terms. The first phrase of introduction is not correct. Third paragraph: 13.15% of total leismaniasis instead of Leishmania. Is necessary a better description of the Figure 1. Also, in the legend, indicating the enlargement of the spleen. Is it necessary to include a image of amastigota identified in case 2. The authors should better explore the conclusion.\n\nIs the background of the cases’ history and progression described in sufficient detail? Partly\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Partly\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Partly\n\nIs the conclusion balanced and justified on the basis of the findings? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-1371
https://f1000research.com/articles/8-1442/v1	15 Aug 19	{ "type": "Study Protocol", "title": "Study protocol for a randomised controlled trial of haloperidol plus promethazine plus chlorpromazine versus haloperidol plus promethazine for rapid tranquilisation for agitated psychiatric patients in the emergency setting (TREC-Lebanon)", "authors": [ "Joseph E. Dib", "Clive E. Adams", "Werner Henry Ikdais", "Elie Atallah", "Hiba Edward Yaacoub", "Tony Jean Merheb", "Francois Kazour", "Fouad Tahan", "Georges Haddad", "Marouan Zoghbi", "Jocelyn Azar", "Chadia Haddad", "Souheil Hallit", "Clive E. Adams", "Werner Henry Ikdais", "Elie Atallah", "Hiba Edward Yaacoub", "Tony Jean Merheb", "Francois Kazour", "Fouad Tahan", "Georges Haddad", "Marouan Zoghbi", "Jocelyn Azar", "Chadia Haddad", "Souheil Hallit" ], "abstract": "Background: Agitated and aggressive behaviours are common in the psychiatric setting and rapid tranquilisation is sometimes unavoidable. A survey of Lebanese practice has shown that an intramuscular haloperidol, promethazine and chlorpromazine combination is a preferred form of treatment but there are no randomised trials of this triple therapy. Methods: This is a pragmatic randomised trial. Setting - the psychiatric wards of the Psychiatric Hospital of the Cross, Jal Eddib, Lebanon. Participants - any adult patient in the hospital who displays an aggressive episode for whom rapid tranquilisation is unavoidable, who has not been randomised before, for whom there are no known contraindications. Randomisation – stratified (by ward) randomisation and concealed in closed opaque envelope by independent parties. Procedure – if the clinical situation arises requiring rapid tranquilisation, medical residents overseeing the patient will open a TREC-Lebanon envelope in which will be notification of which group of treatments should be preferred [Haloperidol + Promethazine + Chlorpromazine (HPC) or Haloperidol + Promethazine (HP)], along with forms for primary, secondary and serious adverse effects. Treatment is not given blindly. Outcome - primary outcome is calm or tranquil at 20 minutes post intervention. Secondary outcomes are calm/tranquil at 40, 60 and 120 minutes post intervention, asleep, adverse effects, use of straitjacket and leaving the ward. Follow-up will be up to two weeks post randomisation. Discussion: Findings from this study will compare the HPC versus HP combination used in Lebanon’s psychiatry emergency routine practice. Trial registration: ClinicalTrials.gov NCT03639558. Registration date, August 21, 2018.", "keywords": [ "Lebanon", "Randomised Controlled Trial", "Rapid Tranquilisation", "Chlorpromazine", "Haloperidol", "Promethazine", "Agitation", "Aggression", "Violence", "Trial Protocol", "Emergency management" ], "content": "Abbreviations\n\nHPC: Haloperidol + Promethazine + Chlorpromazine; HP: Haloperidol + Promethazine; TREC: Abbreviation for the Portuguese ‘Tranquilização Rápida-Ensaio Clínico’ or Rapid Tranquilisation Clinical Trial; RT: Rapid Tranquilisation; ER: Emergency Room; DMC: Data Monitoring Committee; SC: Steering Committee\n\n\nIntroduction\n\nAggressive and violent behaviour is a common behaviour seen in emergency psychiatric presentations with a prevalence of 3–10%1. This aggression is due to a range of psychiatric disorders, such as schizophrenia and bipolar disorder, and/or substance use, personality disorders or dementia2. Guidelines recommend aggressive patients to be ‘verbally tranquilised’ or some form of de-escalation in order for the attending physician to accurately and safely perform a diagnostic history and physical examination3. Aggressive patients make this process difficult or sometimes impossible and carers may be required to work with limited evidence. Since the psychiatric team has a responsibility of ensuring the safety of everyone, rapid and safe tranquilisation may become unavoidable.\n\nRapid tranquillisation (RT) is not a ‘treatment’ but rather a short-term management technique for severely agitated and/or aggressive behaviour in people experiencing severe psychiatric distress. Due to its restrictive nature, RT is a last resort when all other attempts to calm a situation have failed and should always be used in a way that respects human rights and never as a substitute for adequate staffing4.\n\nIn this difficult situation medication(s) are most commonly administered intramuscularly5, and, depending on where in the world the management is happening, physical restraints may include use of a straitjacket6, seclusion room7 or medical restraints8 - binding the patient safely to a bed using two or four points9. Physical restraining by staff to administer medication is common worldwide.\n\nGlobally, guidelines differ in their specific recommendations4,10,11 – often based on the same limited evidence - and then may not be adhered to in local clinical practice12,13. There are no directly relevant national Lebanese guidelines.\n\nIn preparation for this study we surveyed RT practice in the largest (>800 beds) psychiatric hospital in Lebanon (Psychiatric Hospital of the Cross, Beirut). Several different medications were used but the use of a combination of haloperidol plus promethazine plus chlorpromazine (HPC) was common13. Haloperidol plus promethazine (HP) – without the addition of chlorpromazine - was also used. Long clinical experience has proved both combinations effective, but which is best in terms of overall safety remains unclear.\n\nOur systematic searches found that the HPC combination is used elsewhere14 but identified no relevant randomised trials. The HP combination, however, has strong trial-based evidence from low and middle-income countries supporting its use - albeit in comparison with medications other than HPC15. There is an unanswered question as regards the relative effects of HPC versus HP.\n\nThis study takes its name from the first Brazilian TREC study (TREC – abbreviation for the Portuguese ‘Tranquilização Rápida-Ensaio Clínico’ or Rapid Tranquilisation Clinical Trial)16,17 indicating the debt we owe to their pragmatic design and to allow easy identification with an increasing group of related randomised trials.\n\nTREC-Lebanon aims to compare its routine emergency drug HPC versus HP during an agitated episode requiring rapid tranquilisation.\n\nThe SPIRIT 2013 statement which consists of a 33 item checklist of minimum recommended items18. More information can be found on the SPIRIT website (Table 1).\n\nLebanon is a low to middle income country in the Middle-East with a population of approximately 6 million19. The country has three mental hospitals and five psychiatric units within general hospitals (43 psychiatric beds per 10,000; 1 psychiatrist per 100,000)20. The Lebanese Ministry of Health contracts the private sector to provide free treatment for patients who cannot afford to pay (the majority)20. There are no disability benefits for people with mental disorders and no disability funding for mental health21. Psychiatric Hospital of the Cross, set in the metropolitan area of Beirut (>2 million people), provides a service that extends across the whole country.\n\n\nMethods\n\nTREC-Lebanon is a pragmatic trial designed by the main researcher (JD) working in close collaboration with partners in the Psychiatric Hospital of the Cross, Jal Eddib, Lebanon.\n\nThis is the final version of the protocol.\n\nInclusion criteria. An adult (18–64 years) will be eligible if, when presenting at the Psychiatric Hospital of the Cross:\n\ni. he/she requires emergency acute intramuscular medication because of disturbed and dangerous behaviour thought due to psychiatric morbidity; and\n\nii. if the clinician is uncertain of the benefits of the HPC combination over those of HP used together.\n\nExclusion criteria. A person will not be eligible if:\n\ni. The clinician\n\n• knows one treatment regimen has benefit over the other for that particular person;\n\n• is aware of a contra-indication of one of the treatments, such as\n\n■ allergy;\n\n■ past adverse reaction; or\n\n■ already given/taken drugs in the community which would make additional HP or HPC ill-advised;\n\n• does not want to enter the person into the trial for any reason\n\nii. There is an Advanced Directive expressing a wish for one or other, or another treatment in the emergency setting;\n\niii. The person has already been randomised into the trial; and if\n\niv. An accompanying person (friend/family/Police Officer) refuses patient trial entry.\n\nHaloperidol, promethazine and chlorpromazine are included in the WHO's List of Essential Drugs22.\n\nChlorpromazine. Chlorpromazine is a widely used23, effective24 antipsychotic drug, but can cause a number of adverse effects including anticholinergic and antihistaminic effects25. Chlorpromazine is known to be the most epileptogenic of the older antipsychotic drugs causing seizures ranging from 1–4% depending on dosages26.\n\nHaloperidol. Haloperidol is also an older effective27 antipsychotic, less prone than chlorpromazine to cause sedation (less antihistaminic effects) but more causative of movement disorders including acute dystonia (involuntary dramatic contractions of muscles in, for example, the neck, face, pelvis, spinal muscles)28. These acute reactions are not life-threatening but are distressing and frightening to the patient, further eroding trust in the services. Acute dystonia can be swiftly and successfully treated with drugs with anticholinergic/antihistaminic properties, such as promethazine or procyclidine29,30. The occurrence of these reactions, in 1–2% of those given haloperidol alone, was the cause of the early termination of the second TREC trial of Brazil (comparing HP with haloperidol alone)31. The Steering Group of that study felt the evidence was strong enough to make the sole emergency use of intramuscular haloperidol impossible to justify if promethazine was available.\n\nPromethazine. Promethazine hydrochloride is also an old antipsychotic drug but also has potent anticholinergic and antihistaminic properties. This helps offset sickness and movement disorders – including acute dystonia – but also causes it to be sedative32.\n\nThe combinations. Combining drugs can change - increase or decrease - the incidence of known adverse effects33 or result in novel effects unheard of with each drug on its own34,35. The HP combination is well tested, widely used and trusted. Evidence-based guidelines now recommend its use36. We think the triple combination of haloperidol, promethazine and chlorpromazine (HPC) is not so widely employed, but do know it is not only Lebanon that uses it 17. Whether addition of another drug (chlorpromazine) has benefit or causes difficulties may be illustrated by this study, but we identified no existing literature suggesting that there are particular concerns regarding adverse effects of this HPC combination.\n\nThe primary aim of TREC-Lebanon is to investigate whether the proportion of patients calm/tranquil at 20 minutes is any different between the two investigative approaches. In such a stressful situation, a small advantage for an intervention could represent a worthwhile benefit. However, to highlight a clear difference with confidence, this would need larger numbers of people than this phase of work would allow (Table 2). Realistically, taking into account our preliminary work13 and the time constraints of this PhD project with no extramural funding, TREC-Lebanon expects to involve a minimum of 90 patients across a 3-month period. Confidence intervals will be calculated and interpreted according to Altman37. Statistical significance at the 5% level for the primary outcome and at 1% for the secondary outcomes. K statistics will be used for estimating inter-rater agreement for the primary outcome. SPSS version 24 will be used for the analysis of statistical data.\n\nRandomisation will be undertaken in the United Kingdom (by CEA). Small block sizes were chosen to ensure even distribution of treatments (4,6) and whether HPC was to be coded as '1' or '0' was randomly assigned using MS Excel's RAND function (HPC coded as '1'). A free online programme randomised block size, and then treatments to groups within blocks. Using our survey data13, it was thought that 30% of randomised incidents would be on the women's ward - so the first set of complete blocks covering this proportion (therefore up to #32) were taken for the women's ward and men were then started at number 33. Tables of TREC-Envelopes’ number by contents will be constructed and supplied to a Lebanese colleague (SH). The tables will list the contents of the envelopes in groups of ten, not disclosing the block sizes used. SH, always working independently of both the TREC-Lebanon team and CEA, will ensure that the correct medication combination is named within each TREC-envelope before it is sealed.\n\nConcealment of allocation will be ensured by i) not disclosing the randomly varied block sizes to the colleagues packing the envelopes; ii) the supply of tables to those colleagues that gives no suggestion that blocks are even being employed; iii) the independence of those packing the envelopes from the other researchers or the clinicians; and iv) the identical nature of the packed fully opaque envelopes.\n\nThese easy-to-use envelopes will be constructed of cardboard, identical and consecutively numbered. The final check to ensure that nothing has gone wrong with the randomisation will be by the Principal Investigator (CH) filling in a form for each block of ten opened envelopes. At analysis the date and time of instigation of treatment will be ordered and this will order the TREC-IDs showing that each envelope was opened consecutively.\n\nBecause the TREC-Lebanon study evaluates care in the emergency situation, it is imperative that the doctors and nurses know which intervention is being given. The study is blind only up until the time that the TREC-Lebanon trial envelope is opened. Therefore, it is crucial that the evaluation of the severity of a person's disturbance and the first impression on the possible cause for the disturbed behaviour are recorded before the envelope is opened. Once the envelope is opened, doctors and nurses will have knowledge of the medications to be used. It is perfectly feasible that the knowledge that one drug has been given will influence the care beyond the actual effects of the medication. Keeping the study open from that point onwards is not only practical in the emergency situation, but also desirable as the evaluation of care being undertaken is as near real-world circumstances as is possible.\n\nThis trial has been registered and accepted at clinicaltrials.gov: https://clinicaltrials.gov/ct2/show/NCT03639558 (registered prospectively on August 21, 2018).\n\n\nProcedures\n\nAll trial materials, and guidelines for their use, are provided in the TREC-Lebanon folder supplied by the co-ordinating centre. The TREC-Lebanon trial is designed to not interfere with routine care. The eligibility criteria are simple and the process of randomisation fits closely into normal hospital procedures. Data collection will be limited to the minimum necessary and will involve little more than extraction of routine information by a person designated to spend time on the TREC-Lebanon trial. No data are redundant. It is not envisaged that busy doctors and nurses will spend time filling out complicated forms.\n\nBlinding raters would have added additional complexity to the study that would have made the trial much less acceptable to the emergency room (ER) staff. More importantly, it would have completely changed the emphasis of TREC-Lebanon. What is being evaluated is the real-world practice of giving two different drug regimens in the psychiatric emergency setting. In the real-world situation health care professionals know what treatment is being given.\n\nOccasionally patients are given sedative medication by their parents, friends or law officials prior to psychiatric admission13. As far as the hospital clinician is concerned, he or she still decides if the patient should be randomised within the trial if the patient is still exhibiting an agitated episode.\n\nMost people arrive at the hospital’s administrative centre whereby they are registered and then transferred to their ward13. If patients are presenting with violent conduct that could potentially harm people in their vicinity, they are taken directly to their ward while those who brought them in fill out their paperwork – including all necessary consent forms (see section below).\n\nUnlike TREC-Rio17 whereby agitated patients arriving at the psychiatry ER have been given an intervention after the nurse opens the box within the ER, TREC-Lebanon operates differently due to the hospital not having an ER, prompting a different approach. The TREC-Envelopes are placed in the underground research clinic accessible only to researchers and doctors. The envelopes are separated between men and women (stratified randomisation). Residents who are on duty will be granted access to the research director’s clinic where the envelopes are securely kept and will be given two envelopes in a consecutive manner – one for a man and one for a woman while the research assistant keeps track and notes which envelopes have been taken. The resident carries the envelope during duty and in the event of an agitated episode, opens the envelope depending on sex, fills out the form, returns all forms within the envelope and returns it back to the research director’s clinic whereby completed forms will be stored in a separated secured TREC-box.\n\nWhenever possible, carers accompanying the patient should have an opportunity to see the information leaflet (Extended data: TREC-Lebanon information for relatives) before randomisation.\n\nIf the attending doctor decides the person should not be entered into TREC-Lebanon, he/she will notify the resident on duty and anonymised information on this group will be collected:\n\ndate and time\n\nage\n\ngender\n\nethnicity (if applicable)\n\nthe reason not eligible for trial participation, or if they are eligible but failed to be randomised.\n\nIf the attending clinician decides the pateint to be eligible, then the next consecutive envelope is taken from the research director’s clinic, the form on its cover is filled out, and only then is it opened. The trial entry form printed on the sealed envelope (Extended data: TREC-Envelope entry form) records brief baseline details about the person, the severity of disturbance, its presumed cause, the date and time of opening and the name of the doctor. This action constitutes trial entry (Figure 1).\n\nThe residents for the trial are WHI, HEY, EA and TJM. The hospital undergoes frequent resident rotation; therefore, residents undergoing training within the hospital for long durations of time were recruited in order to begin the trial, train incoming residents and see the trial completed. All residents, the Research Director (SH) and the Principle Investigator (CH) are involved in a WhatsApp group chat to inform, keep track and ask questions should they arise during the course of the trial.\n\nEvery envelope has the entry form printed on it. Randomisation proceeds using a local pack system with the trial team providing the identical sealed envelopes made of sturdy fully opaque card.\n\nEach envelope contains:\n\nPaper slip indicating the use of Haloperidol + Promethazine (HP)\n\n1 × TREC-Lebanon follow-up form (Extended data: Primary measures of outcome form)\n\n2 × TREC-Lebanon stickers for the drug prescription form and medical notes\n\nOr\n\nPaper slip indicating the use of Haloperidol + Promethazine + Chlorpromazine (HPC)\n\n1 × TREC-Lebanon follow-up forms (Extended data: Primary measures of outcome form, Dr. Stopwatch form and TREC-Lebanon main data collection form)\n\n2 × TREC-Lebanon stickers for the drug prescription form and medical notes\n\nAll doses used are at the discretion of the attending clinician. If the contents of a trial envelope are destroyed, or unfit for use, the person should not be randomised a second time and the equivalent material should be obtained from the usual hospital supplies.\n\nIn the event of continuing aggression despite the TREC-Lebanon medication, ongoing emergency management is entirely at the discretion of the clinicians; another envelope is not opened.\n\nIt is crucial that follow-up is complete and accurate for everyone entered into the study. As a pragmatic study, causing minimal interference with routine care, TREC will not employ any rating scale outcomes. It is likely that completion of scales would be inaccurate and incomplete, validity and reliability would be in question and clinical utility problematic. The main outcome of TREC-Lebanon is tranquillisation by 20 minutes. This was the primary outcome requested by the nursing and medical staff of the hospital13. By asking the relevant clinical staff to select the primary outcome for TREC-Lebanon we hoped to ensure maximum compliance with the trial protocol. Therefore, upon injection of the patient, a timer is started on the resident’s phone and this rings at 20 minutes and then again at 40, 60 and 120 minutes. At each period the attending resident rates whether the person, in their opinion, is tranquil, asleep, has shown adverse effects or needs additional treatment and records this on the follow-up form found originally within the sealed envelope (Extended data: Primary measures of outcome form). The person is considered tranquillised when they are felt to be calm and peaceful, but not asleep. They should not be agitated or restless, nor displaying threatening verbal behaviour, physical aggression against objects, self-aggression or physical aggression to other people.\n\nThe Dr. Stopwatch form (Extended data) is included within the envelope and contains two boxes: time the resident felt the patient was tranquilised either tranquil or asleep and a comments box. The tranquilised box is the ‘true tranquilisation’ outside the 20, 40, 60 and 2 hour time intervals within the primary measure of outcome form, assuming the patient becomes tranquilised after the 2 hour maximum limit. The additional comments box is included if the resident needs to note additional information (i.e. why the patient did not calm down within the 2 hour time frame). The resident starts their own personal stopwatch from the time of intervention to when he/she feels the patient has become calm/tranquil or asleep.\n\nAdditional data are recorded at 24 hours and finally at two weeks (Extended data: TREC-Lebanon main data collection form). These additional data are to be extracted from routine notes. If the patient is transferred to another hospital, the co-ordinating centre will contact every relevant hospital to find out further details on what happened after transfer.\n\nAll data for TREC-Lebanon will be transcribed and collated from the TREC-envelope forms, the follow-up form, severe adverse event form and routine notes of each ER or ward into a form within MS Access – a database management system from Microsoft (Extended data: TREC-Lebanon main data collection form). These anonymised data, in compliance with the ethics committee requests, will be protected as the hard copy forms do not leave the research director’s clinic. All transcribed raw data used by the main researcher (JD) will have personal information such as names of patients abbreviated to retain anonymity.\n\nAnalysis will take place using Statistical Package for Social Sciences (SPSS)38. This will take the form of simple frequencies – to test the integrity of the data, and, for binary outcomes, relative risks and respective 95% confidence intervals, and for continuous data mean differences and their 95% confidence intervals. Tables for this analysis are prepared before recruitment of the first patient (Extended data: Dummy statistical tables). No additional analyses are anticipated. All analysis will be based on groups as randomly allocated; this will be an intention-to-treat analysis.\n\nAll data that includes private information will be anonymised. Once the study is completed, access to raw data containing personal patient information will only be accessible to the hospitals’ director and responsible clinician.\n\n\nAnticipated risks\n\nIt is possible though highly improbable that patients who do not fit the trial’s entry criteria may enter the study. Those that do will not be counted as part of the trial and their notes will be disregarded. Detecting ineligible patients will be seen in the data entry form left in a bin on each ward (the TREC-bin into which all filled out envelopes are put) making it simple and direct to trace.\n\nAttending resident and nurse should monitor action of given intervention, i.e. make sure given treatment has been injected properly. In the event that an agitated patient may break any of the treatment tools – mainly the syringe containing the treatment intervention or destroy the vial containing the intervention before being placed within the syringe, another TREC envelope should not be opened. Instead, attending resident should carry on as per hospital protocol and fill out the serious event form detailing the circumstances (Extended data: Serious event form). Nurses should also detail the nature of compliance as they would normally do in their notes. In the event a patient does break the syringe or capsule, the situation is rectified with the nurse bringing the emergency treatment as detailed in the paper slip in the envelope.\n\nA feasibility phase will take place before the trial commences. The feasibility phase will include a limited number of envelopes (i.e. 5) with contents known to the trialists. The feasibility phase is designed to test the trial’s procedure in practice in order to assess if any unforeseen circumstances arise. In the event such unforeseen circumstances do arise, the trialists will rectify and mediate in the most practical way possible. Changes will be noted by the main researcher (JD) and updated in the trial protocol.\n\nAfter trial entry, clinical events are recorded, as usual, in the patients' notes. Complications and adverse events should be managed as usual. A serious unexpected event form (Extended data: Serious event form) is provided and will be sent to the TREC-Lebanon Co-ordinator (JD) as soon as it is completed.\n\n\nEthics approval and consent to participate\n\nThe Helsinki Declaration39, the European Directive on Clinical Trials40 and the Nuffield Council documents on bioethics41 state that trials in non-consenting patients are permitted on two conditions: i) no other context exists in which to answer the question; and ii) all trial participants receive clear therapeutic benefit from whichever arm they are randomised to. Consideration also has to be given to the local legislation21, namely:\n\n1. Lebanese Act no. 72-9/9/1984 Welfare Act and Protection and Treatment of Mentally Ill Patients;\n\n2. Lebanese Act no. 673-16/3/1998 Narcotic Drugs and Psychotropic Substances and Precursors;\n\n3. Lebanese Act no. 220-29/5/2000 Rights of Mentally Handicapped in Lebanon; and\n\n4. Lebanese Act no 574-11/2/2004 Patients’ Rights and Informed Consent;\n\nwhich also protects the rights of patients and their families and carers.\n\nAggressive patients in a situation of psychiatric emergency are not able to give consent for their participation in a study42 as their aggressive state, along with their psychiatric disorder places them in a state whereby they lack the capacity to do so, e.g. a patient with schizophrenia suffering from hallucinations is in a state where he or she is not mentally competent to understand the conditions of the treatments involved43. In the case where parents or legal guardians are unavailable, a third party must confirm to the best interest standard whereby the decision made is the most beneficial to the patient. This is a direct application of the principle of beneficence and proportionality: maximize benefit and avoidance of harm. The substituted judgment standard aims to implement the subjective preferences of the patient44.\n\nIn routine care medication is usually given against the will of the patient. Therefore, for TREC-Lebanon, in the same way that doctors are responsible for the choice of a treatment in routine care, they take responsibility for the recruitment of a patient into the study. TREC-Lebanon will not involve administering an inactive compound to those who clearly need sedation/tranquillisation. Both treatments calm aggressive disturbed people13, so there is no 'experimental' intervention. What is still uncertain is the speed for the onset of action, the duration of the effects and the different kinds of adverse reactions. TREC-Lebanon will attempt to answer clinical questions to help the care of this group of people.\n\nA patient/carer information leaflet about TREC-Lebanon is available for all for whom a TREC-Lebanon envelope is opened (in English and Lebanese Arabic). Carers will always be free to decide that their relative should not be entered. Not being involved in TREC-Lebanon will not affect the person's standard of care. An information sheet is provided detailing the aim and purpose of the study (Extended data: TREC-Lebanon information for relatives).\n\nPatients who do not have parents or legal guardians will still be randomised if they are aggressive and fit the inclusion criteria, but will be given a patient consent form post randomisation detailing the background of the study and their right to withdraw all personal data from the trial for whatever reason they see fit (Extended data: TREC-Lebanon consent form for patients)\n\nEthics Application File from the Hospital of the Cross was filled out by the trial coordinator detailing the background and procedure of the trial (non-consenting nature, data sharing agreement, liability waiver) were read and signed before gaining ethical approval and IRB letter.\n\nTREC-Lebanon did not require ethics approval from the University of Nottingham (Ethics number: 271) and has passed ethics approval from the Psychiatric Hospital of the Cross, Beirut, Lebanon (Ethics number: HPC 001/2018).\n\nThe Declaration of Helsinki45 requires the informed consent of participants in randomized controlled trials (paragraph 25). However, according to paragraph 28, if the individual for whatever reason is unable to provide consent, a legally authorised individual – usually a family member or guardian may provide on behalf. Finally, paragraph 30 states if no family members are present and the research is unable to be delayed, randomisation may proceed provided the specific reasons to why the patient is unable to provide consent (in this case, mental illness with violent display) and the research has been approved by an ethics committee. TREC-Lebanon abides by the Helsinki Declaration and provides a consent form to all patients who were randomised without having the ability to consent at the time or a family member/guardian to consent on their behalf.\n\n\nTrial Organisation\n\nThe TREC-Lebanon Co-ordinating Group: The co-ordinating centre of the Lebanese arm is based at the Institute of Mental Health, University of Nottingham, United Kingdom. The Co-ordinating Group has overall responsibility for the design of the proposed trial and is responsible for all aspects of day-to-day trial administration. The Co-ordinating Group is also responsible for preparing reports for the Steering Committee. Membership: JD, CEA, SH.\n\nThe overall progress of the trial, adherence to protocol, patient safety and the consideration of new information will be monitored by a scientific and administrative Steering Committee (SC). At the end of the proposed study period, the SC will consider the extension of the study, to allow the detection of other important effects. Membership: PS and RH.\n\nTREC-Lebanon will include a committee to oversee progress of the trial. Since TREC-Lebanon might take three to six months to complete, an independent Data Monitoring Committee (DMC) will, in confidence, monitor results. This could be undertaken on a week to week or month to month basis depending on the collective agreement of all the members of the DMC. In the light of the interim data, and of any other evidence or advice they wish to seek, the DMC will inform the chair of the SC if, in their view: i) there is proof beyond reasonable doubt that for any particular group or subgroup treatment with one or other regiment is clearly indicated or contraindicated; or, ii) it is evident that no clear outcome will be obtained. Proof beyond reasonable doubt may be taken as the difference of at least three standard deviations and at least one of the primary outcomes.\n\nThe DMC may communicate certain interim analysis to the SC or suggest certain protocol changes, but the SC will remain responsible for deciding which changes to adopt. Membership: GA and JM. The committee will receive the first batch of data when trial participants are at a total of 50 along with information such as adverse effects, unforeseen circumstances and trial progress so far.\n\n\nCurrent Study Status\n\nAt the time of submission of this manuscript, trial was still ongoing albeit approaching its final phase. Presently, trial has concluded recruitment process. Data has been transcribed and awaiting analysis.\n\n\nDissemination of Study Outcomes\n\nThe results of TREC-Lebanon will be published in at least one peer reviewed indexed journal and will be presented in relevant conferences.\n\n\nDiscussion\n\nAs mentioned earlier, violence in the psychiatry setting is common and rapid tranquilisation is sometimes necessary and unavoidable. Not only are national guideline recommendations limited in their evidence backing - but surveys of practice have shown to differ from clinicians’ opinions on management during an agitated episode13.\n\nThe HP and HPC combination have both been used in routine care outside Lebanese practice16, but as far as systematic searching has shown, there exists no clinical trials randomising the HPC combination, making it an ideal candidate for randomisation.\n\nPossible limitations for this trial are that there are no ERs; therefore residents must always carry the TREC envelopes at all times, increasing the risk for error (i.e. misplacing envelope, using wrong envelope, etc.). Despite the limitation, the chances of error remain low due to the small sample size of 100.\n\nOverall, since TREC-Lebanon is comparing two interventions with drug combinations that are used in its routine practice13, we assume in both cases that rapid tranquilisation will be achieved reducing the risk of the agitated patients harming themselves, as well as harming others.\n\n\nData availability\n\nNo underlying data is associated with this article.\n\nOpen Science Framework: TREC-Lebanon Protocol, https://doi.org/10.17605/OSF.IO/MYCQ946.\n\nThis project contains the following extended data:\n\n- TREC-Lebanon information for relatives\n\n- TREC-Envelope entry form\n\n- Primary measures of outcome form\n\n- Dr. Stopwatch form\n\n- TREC-Lebanon main data collection form\n\n- Dummy statistical tables\n\n- Serious event form\n\n- TREC-Lebanon consent form for patients\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).", "appendix": "Author contributions\n\n\n\nJD, under the supervision of CEA, co-conceptualised and designed the study. SH and CH overlooked study setting, including packaging, labelling and storing all TREC forms. Residents WHI, HEY, EA and TJM, under the supervision of their attending doctors (FT, FK, GH, MZ and JA) volunteered to be part of the TREC-Lebanon trial, and are the residents that will administer the intervention and fill out all TREC forms.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nThe authors acknowledge the efforts of all employees at the Psychiatric Hospital of the Cross, including all researchers, physicians, residents and nurses whose partaking in the trial make it possible.\n\nThe authors acknowledge all the efforts of the University of Nottingham and at the Institute of Mental Health Facility where this study was planned and passed ethics approval.\n\n\nReferences\n\nTardiff K, Koenigsberg HW: Assaultive behavior among psychiatric outpatients. Am J Psychiatry. 1985; 142(8): 960–3. PubMed Abstract \| Publisher Full Text\n\nKaplan HI, Sadock BJ: Kaplan and Sadock's synopsis of psychiatry: Behavioral sciences/clinical psychiatry. Williams & Wilkins Co; 1998. Reference Source\n\nTreatment of schizophrenia 1999. The expert consensus guideline series. J Clin Psychiatry. 1999; 60 Suppl 11: 3–80. PubMed Abstract\n\nNational Institute for Clinical Excellence: Short-term Management of Violent (Disturbed) Behaviour in Adult Psychiatric In-patient and Accident and Emergency Settings. Second Draft for Consultation. London: NICE; 2004. Reference Source\n\nYildiz A, Sachs GS, Turgay A: Pharmacological management of agitation in emergency settings. Emerg Med J. 2003; 20(4): 339–346. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nColaizzi J: Seclusion & restraint: A historical perspective. J Psychosoc Nurs Ment Health Serv. 2016; 43(2): 31–37. PubMed Abstract \| Publisher Full Text\n\nCrespi TD: Restraint and seclusion with institutionalized adolescents. Adolescence. 1990; 25(100): 825–829. PubMed Abstract\n\nFisher WA: Restraint and seclusion: a review of the literature. Am J Psychiatry. 1994; 151(11): 1584–1591. PubMed Abstract \| Publisher Full Text\n\nSaks ER: The Use of Mechanical Restraints in Psychiatric-Hospitals. Yale Law J. 1986; 95(8): 1836–1856. Reference Source\n\nAllen MH, Currier GW, Hughes DH, et al.: The expert consensus guideline series. Treatment of behavioral emergencies. Postgrad Med. 2001; (Spec No): 1–88; quiz 89-90. PubMed Abstract\n\nWing J, Marriott S, Palmer C, et al.: Management of imminent violence: clinical practice guidelines to support mental health services. 1998. Reference Source\n\nPilowsky LS, Ring H, Shine PJ, et al.: Rapid tranquillisation. A survey of emergency prescribing in a general psychiatric hospital. Br J Psychiatry. 1992; 160: 831–835. PubMed Abstract \| Publisher Full Text\n\nDib JE, Adams CE, Kazour F, et al.: Managing acutely aggressive or agitated people in a psychiatric setting: a survey in Lebanon. Med J Islam Repub Iran. 2018; 32: 60. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHuf G, da Silva Freire Coutinho E, Fagundes HM Jr, et al.: Current practices in managing acutely disturbed patients at three hospitals in Rio de Janeiro-Brazil: a prevalence study. BMC Psychiatry. 2002; 2: 4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHuf G, Alexander J, Gandhi P, et al.: Haloperidol plus promethazine for psychosis-induced aggression. Cochrane Database Syst Rev. 2016; 11: CD005146. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTREC Collaborative Group: Rapid tranquillisation for agitated patients in emergency psychiatric rooms: a randomised trial of midazolam versus haloperidol plus promethazine. BMJ. 2003; 327(7417): 708–13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHuf G, Coutinho ES, Adams CE, et al.: TREC-Rio trial: a randomised controlled trial for rapid tranquillisation for agitated patients in emergency psychiatric rooms [ISRCTN44153243]. BMC Psychiatry. 2002; 2(1): 11. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChan AW, Tetzlaff JM, Altman DG, et al.: SPIRIT 2013 statement: defining standard protocol items for clinical trials. Ann Intern Med. 2013; 158(3): 200–207. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEl Laithy H, Abu-Ismail K, Hamdan K: Poverty, growth and income distribution in Lebanon. In. International Policy Centre for Inclusive Growth: International Policy Centre for Inclusive Growth; 2008. Reference Source\n\nWorld Health Organization: WHO - Aims report on mental health system in Lebanon. Beirut, Lebanon Ministry of Health Lebanon. 2010. Reference Source\n\nBerger EL: Mental Health Atlas 2005. J Nerv Ment Dis. 2007; 195(4): 358–358. Reference Source\n\nWorld Health Organization: The Selection and Use of Essential Medicines: Report of the WHO Expert Committee, 2015 (including the 19th WHO Model List of Essential Medicines and the 5th WHO Model List of Essential Medicines for Children). World Health Organization; 2015.\n\nAhmed U, Jones H, Adams CE: Chlorpromazine for psychosis induced aggression or agitation. Cochrane Database Syst Rev. 2010; (4): CD007445. PubMed Abstract \| Publisher Full Text\n\nAdams CE, Awad GA, Rathbone J, et al.: Chlorpromazine versus placebo for schizophrenia. Cochrane Database Syst Rev. 2014; (1): CD000284. PubMed Abstract \| Publisher Full Text\n\nAbidi S, Bhaskara SM: From chlorpromazine to clozapine--antipsychotic adverse effects and the clinician's dilemma. Can J Psychiatry. 2003; 48(11): 749–755. PubMed Abstract \| Publisher Full Text\n\nHedges D, Jeppson K, Whitehead P: Antipsychotic medication and seizures: a review. Drugs Today (Barc). 2003; 39(7): 551–557. PubMed Abstract \| Publisher Full Text\n\nDold M, Samara MT, Li C, et al.: Haloperidol versus first-generation antipsychotics for the treatment of schizophrenia and other psychotic disorders. Cochrane Database Syst Rev. 2015; 1(1): CD009831. PubMed Abstract \| Publisher Full Text\n\nRasmussen SA, Rosebush PI, Mazurek MF: The Relationship Between Early Haloperidol Response and Associated Extrapyramidal Side Effects. J Clin Psychopharmacol. 2017; 37(1): 8–12. PubMed Abstract \| Publisher Full Text\n\nAdams CE, Bergman H, Irving CB, et al.: Haloperidol versus placebo for schizophrenia. Cochrane Database Syst Rev. 2013; (11): CD003082. PubMed Abstract \| Publisher Full Text\n\nvan Harten PN, Hoek HW, Kahn RS: Acute dystonia induced by drug treatment. BMJ. 1999; 319(7210): 623–626. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHuf G, Coutinho ES, Adams CE, et al.: Rapid tranquillisation in psychiatric emergency settings in Brazil: pragmatic randomised controlled trial of intramuscular haloperidol versus intramuscular haloperidol plus promethazine. BMJ. 2007; 335(7625): 869. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCantisani C, Ricci S, Grieco T, et al.: Topical promethazine side effects: our experience and review of the literature. Biomed Res Int. 2013; 2013: 151509. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLemmens P, Brecher M, Van Baelen B: A combined analysis of double-blind studies with risperidone vs. placebo and other antipsychotic agents: factors associated with extrapyramidal symptoms. Acta Psychiatr Scand. 1999; 99(3): 160–170. PubMed Abstract \| Publisher Full Text\n\nRummel-Kluge C, Komossa K, Schwarz S, et al.: Second-generation antipsychotic drugs and extrapyramidal side effects: a systematic review and meta-analysis of head-to-head comparisons. Schizophr Bull. 2012; 38(1): 167–177. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAlomar MJ: Factors affecting the development of adverse drug reactions (Review article). Saudi Pharm J. 2014; 22(2): 83–94. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNational Institute for Health Care Excellence: Violence and Aggression: Short-Term Management in Mental Health, Health and Community Settings: Updated edition. 2015. PubMed Abstract\n\nAltman DG: Confidence intervals for the number needed to treat. BMJ. 1998; 317(7168): 1309–1312. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSPSS MOD: SPSS (Statistical Package for the Social Sciens). 2015.\n\nWorld Medical Association: World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA. 2013; 310(20): 2191–2194. PubMed Abstract \| Publisher Full Text\n\nEuropean Parliament: Directive 2001/20/EC of the European Parliament and of the Council of 4 April 2001 on the approximation of the laws, regulations and administrative provisions of the member states relating to the implementation of good clinical practice in the conduct of clinical trials on medicinal products for human use. Med Etika Bioet. 2002; 9(1–2): 12–19. Reference Source\n\nNuffield Council: Nuffield Council on Bioethics. London; 2000.\n\nRebers S, Aaronson NK, van Leeuwen FE, et al.: Exceptions to the rule of informed consent for research with an intervention. BMC Med Ethics. 2016; 17: 9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLesage P, Portenoy RK: Ethical challenges in the care of patients with serious illness. Pain Med. 2001; 2(2): 121–130. PubMed Abstract \| Publisher Full Text\n\nRoth LH, Appelbaum PS, Lidz CW, et al.: Informed consent in psychiatric research. Rutgers Law Rev. 1987; 39(2–3): 425–441. PubMed Abstract\n\nHelsinki TDo: Ethical principles for medical research involving human subjects. 2013.\n\nDib J: TREC-Lebanon Protocol. OSF. 2019. http://www.doi.org/10.17605/OSF.IO/MYCQ9" }	[ { "id": "55715", "date": "05 Nov 2019", "name": "Sophie Hirsch", "expertise": [ "Reviewer Expertise Violence and coercion in psychiatry", "targeted therapies in neurooncology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a very important field of research, due to the fact that randomized controlled trials on people with severe mental illness showing aggressive behavior is still scarce and data for improving safety of this patient group is urgently needed. There are many positive points in the study design that should be highlighted: First, the opportunity for highly disturbed patients to give/withdraw informed consent for study participation after randomization to ensure that these patients are not excluded systematically and the organizational efforts that are made to keep the study recruiting even during night/weekend duty. Second, the blinded rating before the administration of the drug. Third, the 2 hours and 24 hours follow-ups which are missing in some of the rapid tranquilization studies guidelines have to rely on. And last but not least, the consideration of gender issues and the fact that patients who are not eligible for the study are also recorded.\nFor this is a study protocol, it would have been better, to have it published before starting the recruitment process. Due to the fact, that the recruitment process is already completed comments or suggestions for improvement from the reviewers cannot be taken into account anymore. I would like to name some of these flags anyway: The evidence for the combination of haloperidole and promethazine is already quite good (evidence level 1a). Despite this high quality evidence, guidelines at least in Europe are constrained in recommending this combination due to safety reasons (especially QTc prolongation). At the moment, clinicians face the situation that a combination that is favored by international evidence and at least some clinical practice guidelines is not approved by the agencies. Adding a third active ingredient makes the adverse events and interdependencies even more unclear (e. g. chlorpromazine plus haloperidole and the risk for seizures). Against the background of this discussion and the recent discussion if the tranquilizing effect of a drug combination could also be achieved by a higher dose of one drug as monotherapy, the study design with a two- and a three-drug-combination and even the doses of these combinations left at the discretion of the clinician seems questionable. Especially, when rating the effects of the drugs at the follow-up points dosages and additional drugs should be analyzable per patient. It stays unclear if these longitudinal assessments are possible within the scope of the study. One sentence explaining the SPIRIT checklist would have been helpful.\nConcerning protection of data privacy, it would be interesting how the abbreviations for the patients were built up and which data was exchanged via the Whatsapp group.\n(Note to my \"partly\" answer on the last question: I missed the sheets that had to be filled in by the doctors. I did not understand why the form \"Dr. Stopwatch\" and \"TREC-Lebanon main data collection form\" only were present in the HPC- and not in the HP-envelope.)\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Partly", "responses": [ { "c_id": "5520", "date": "19 May 2020", "name": "Joseph Dib", "role": "Author Response", "response": "Hello,On behalf of myself and the TREC-Lebanon team, we would like to thank you for taking the time and responding with valuable input pertaining to the protocol-manuscript.To address your points: while it is true that it would have been better to have the protocol published before recruitment began, the authors (including myself) were on a limited time schedule and had to do with what time we were given. For instance, the resident doctors that were part of the trial team i.e. Administering the intervention, etc would have completed their rotation and left the hospital which would have compromised the study.Adding a third active ingredient makes the adverse events and interdependencies even more unclear (e. g. chlorpromazine plus haloperidole and the risk for seizures). Against the background of this discussion and the recent discussion if the tranquilizing effect of a drug combination could also be achieved by a higher dose of one drug as monotherapy, the study design with a two- and a three-drug-combination and even the doses of these combinations left at the discretion of the clinician seems questionable. The trial did address the adverse events pertaining to adding a third active ingredient to the haloperidol plus promethazine combination. The protocol does indeed state the doses were at the discretion of the clinicians. However, during the trial itself, the doses were fixed at 5mg Haloperidol, 25mg promethazine and 100mg chlorpromazine.Concerning protection of data privacy, it would be interesting how the abbreviations for the patients were built up and which data was exchanged via the Whatsapp group. The WhatsApp group allowed updates to trial recruitment as well as open questions pertaining to the trial. Patient names were not used nor were they required since trial numbers were allocated. For instance, if a patient was admitted and randomised, the resident in charge would message the group with \"New patient number 14, female\". At this point, the trial co-ordinator (Myself) would note the date on the group. This was to keep track on trial recruitment.(Note to my \"partly\" answer on the last question: I missed the sheets that had to be filled in by the doctors. I did not understand why the form \"Dr. Stopwatch\" and \"TREC-Lebanon main data collection form\" only were present in the HPC- and not in the HP-envelope.) We apologise if this was not as clarified as thought. Both the Dr. Stopwatch form and the main data collection form was present in both the HPC and HP envelope.Once again, we thank you for your time and efforts and hope we have clarified the best we can. We look forward to sharing the trial itself when published.Regards,Joseph Dib" } ] }, { "id": "63552", "date": "19 May 2020", "name": "Stefan Leucht", "expertise": [ "Reviewer Expertise Meta-analysis of treatments for schizophrenia", "clinical trials." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors submit the protocol of the clinical trial comparing to combinations of antipsychotics for emergency patients which need a rapid intervention. The protocol is a very detailed and explains every important step of the study. The team around the study, in particular the author Prof Clive Adams, has a lot of experience in this type of study. He has helped to contact and published several trials of this kind in the past. The current trial will fill another bit of the puzzle about the best intervention for acutely ill mentally ill patients who need rapid treatment. As the interventions are frequently used in the trial centre so that the clinicians have a lot of experience with them, I do also not see an important risk going beyond routine care.\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Not applicable", "responses": [] }, { "id": "63478", "date": "29 May 2020", "name": "Gisele Huf", "expertise": [ "Reviewer Expertise Clinical epidemiology" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript reports a protocol for a randomised clinical trial to evaluate the best pharmacological management of agitated patients in psychiatric emergency rooms in Lebanon. It follows a previous version of this protocol, published in 2002, for studies conducted in Brazil and India, assessing the use of the prevalent intervention in both countries, a combination of haloperidol and promethazine. The Lebanon study will compare this intervention with their local practice: a combination of haloperidol, promethazine and chlorpromazine. The similarities with the original protocol are many. The pragmatic design, the blinding up to the point of randomisation just to ensure concealment of allocation, the type of participants, the doses of medication at the discretion of the doctor, the main and secondary outcomes, and the analysis of data. The main differences are in the randomisation process: it is stratified by ward and delivered in sealed envelopes instead of sealed boxes containing the treatment. Although the trial intends to recruit only 90 participants, the results of this study could be added to the body of evidence in systematic reviews. I have only some minor comments:\nThe authors state they will perform an intention-to-treat analysis, but they also declare that if ineligible people enter the trial, they will not be counted and their notes will be disregarded. Isn’t it a contradiction?\n\nThe authors state that in the event that some syringe or vial containing medication is broken, another envelope should not be opened. It doesn´t make much sense.\n\nThe authors state that guidelines are “often based on the same limited evidence”. This is currently not accurate and the references used (1998, 2001 and 2004) are outdated.\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1442
https://f1000research.com/articles/7-1798/v1	15 Nov 18	{ "type": "Research Article", "title": "Anti-hypercholesterolemic effect of Zingiber montanum extract", "authors": [ "Swandari Paramita", "Meiliati Aminyoto", "Sjarif Ismail", "Enos Tangke Arung", "Swandari Paramita", "Meiliati Aminyoto", "Sjarif Ismail" ], "abstract": "Background: Hypercholesterolemia, high cholesterol levels in the blood, can contribute to many forms of disease, most notably cardiovascular disease. Anti-hypercholesterolemic agents generally used for those conditions have several side effects for patients. Zingiber montanum, known locally as “bangle”, belongs to the family Zingiberaceae and is a potential plants for alternative anti-hypercholesterolemic agents. This plant, from East Kalimantan, is used in traditional medicine for health problems caused by high cholesterol levels. The aim of this research was to find alternatives to anti-hypercholesterolemic agents, especially from natural sources. Methods: This study was an experimental study using 30 Wistar male white rats. Subjects were randomly divided into 6 groups (n=5): (1) normal control group; (2) high fat diet control group; (3) high fat diet with simvastatin; (4-6) high fat diet with Zingiber montanum extracts 100, 200, and 400 mg/kg. After 4 weeks of treatment, blood was collected from all groups, and plasma concentrations of triglycerides, total cholesterol, high density lipoproteins (HDL), and low density lipoproteins (LDL) were measured. Results: The results showed significant differences in total cholesterol (p=0.000), LDL (p=0.000) and triglycerides (p=0.001) in the high-fat diet group with Z. montanum extract, as compared to the high-fat diet control. Meanwhile, there were no significant differences in HDL levels (p=0.830) between the high-fat diet group and other groups. The results also showed significant differences in total cholesterol and LDLs for rats treated with Z. montanum extract, 100 mg/kg (p=0.000), 200 mg/kg (p=0.000), and 400 mg/kg (p=0.000) compared to the high-fat diet group. The result of Z. montanum 400 mg/kg also showed a significant reduction, not only for total cholesterol and LDLs, but also for triglycerides (p=0.030). Conclusion: It could be concluded that Z. montanum extracts have the potency to be further developed as a new natural source of the anti-hypercholesterolemic agents.", "keywords": [ "anti-hypercholesterolemic", "Zingiber montanum" ], "content": "Introduction\n\nHypercholesterolemia is a condition characterized by very high levels of cholesterol in the blood1. Excess cholesterol in the bloodstream can be deposited into the walls of blood vessels. Hypercholesterolemia certainly predicts coronary heart disease risk2. Numerous agents can be used for hypercholesterolemia patients, one of them is HMG CoA reductase inhibitors or statins (i.e. Simvastatin)3. Side effects of statins are usually very well tolerated but can cause hepatitis-like symptoms and myopathy4.\n\nMany plants have been used in traditional medicine in Indonesia. Zingiber montanum (J.Koenig) Link ex A.Dietr. (in Indonesian is called Bangle) are potent as antihypercholesterolemic5. Among the synonyms, Zingiber cassumunar Roxb. and Zingiber purpureum Roscoe are commonly used for Z. montanum6. In Southeast Asian countries, Z. montanum is well-known for its anti-inflammatory properties7. Z. montanum is used as a traditional medicine in East Kalimantan for health problems caused by high cholesterol level8. The aim of the present study was to evaluate the anti-hypercholesterolaemic effect of Z. montanum in rat models of hypercholesterolemia.\n\n\nMethods\n\nThe sampling of medicinal plants was conducted in the Kutai Kartanegara District, East Kalimantan (0°24’18.4”S 117°4’24.7”E).\n\nThe rhizomes of Z. montanum were sliced and dried at room temperature for 3 days, crushed and transferred into a glass container. Crushed rhizomes was soaked in absolute ethanol (9401-03 Alcohol, Anhydrous, Reagent, J.T. Baker) for 5 days. The mixture was shaken occasionally with a shaker (3525 Incubator Orbital Shaker, Lab-Line, US). After 5 days, the materials were filtered (Whatman Filter Paper 11µm, Sigma-Aldrich) and evaporated using a rotary evaporator (RV06-ML Rotary Evaporator, IKA, Germany). The dried extracts were obtained and stored at 4°C in a dark bottle until use.\n\nBased on Federer’s rule, with six group of induction, 30 male Wistar rats (Rattus norvegicus), weighing 250–350g, aged 12–13 months, were obtained from Animal House Faculty of Medicine (Mulawarman University) and randomly divided into 6 groups: control, high fat diet, high fat diet with simvastatin and high fat diet with 3 different doses of Z. montanum extract (100, 200, and 400 mg/kg). They were acclimatized for 1 week in a controlled room temperature of 25°C, with a 12-hour light/dark cycle, and access to food pellets and filtered water ad libitum to adapt to the new environment. They were housed in wire cages (30×30×30 cm), one animal in each cage. All the treatment rats were given high fat diet for 4 weeks with 10% chicken egg yolk and reused cooking oil to their standard pellet diet (Japfa, Comfeed, Indonesia) with tap water ad libitum.\n\nAfter 4 weeks of treatment, blood was collected from all groups after an overnight fasting. All animals were anesthetized intraperitoneally with a ketamine injection (Hameln, Germany) at a dose of 60 mg/kg before blood was taken. After anesthetized, animals were euthanized by cervical dislocation. Blood was aspirated through the left ventricle of each animals’ heart. Two mililiters of blood was aspirated using a 3 ml disposable syringe and then inserted in a vaccutainer tube with an anticoagulant. Plasma concentrations of triglycerides, total cholesterol, high density lipoproteins (HDLs), and low density lipoproteins (LDLs) were measured with an automatic analyzer system (BiOLis 24i; Boeki, Tokyo, Japan).\n\nAll statistical analysis was performed using SPSS version 16.0 for Windows. Data normality was examined using Shapiro-Wilk normality test. Then data were analyzed using ANOVA and post hoc with Tukey test. A p value of ≤ 0.05 was considered to be significant.\n\nAll protocols used in this experiment received approval from the Ethical Animal Care from the Medical and Health Research Ethics Commission, Faculty of Medicine, Mulawarman University with approval number 81/KEPK-FK/V/2018. All efforts were made to ameliorate any suffering of animals used in this research.\n\n\nResults\n\nThe results showed that significant differences between total cholesterol (p=0.000), LDL (p=0.000) and triglycerides (p=0.001) (Figure 1) levels achieved between high fat diet group and Z. montanum extracts. Meanwhile, there were no significant differences in the HDL (p=0.830) level between the high fat diet group and other groups. Tukey post hoc test showed significant differences between total cholesterol (p=0.000) and LDL (p=0.000) levels with the high fat diet group. The results of Z. montanum 400 mg/kg also showed a significant reduction, not only for total cholesterol, but also for triglyceride (p=0.030) levels (Table 1).\n\nNote: HFD = high-fat diet, SIM = simvastatin; ZM-1 = Z. montanum 100 mg/kg;\n\nZM-2 = Z. montanum 200 mg/kg; ZM-3 = Z. montanum 400 mg/kg\n\n*Tukey post hoc test significant p<0.05 compared to HFD control\n\n\nDiscussion\n\nZ. montanum (Supplementary File 1) is used medicinally in Asia, primarily as a carminative and stimulant for the stomach, and to treat diarrhea and colic9. Pharmacological properties of Z. montanum include antimicrobial activity, anti-oxidant activity, insecticidal-activity, anti-cancer, anticholinesterase activity, and anti-inflammatory10. The main constituents, terpinen-4-ol and DMPBD, has been found to be effective against bacteria and also have anti-inflammatory activity11.\n\nThe rhizome extracts of Z. montanum showed the highest total curcuminoid content compared to other species of Zingiber7,12. Curcuminoid isolated from Z. montanum may possess a potent protective action against oxidative stress13. The major target for anti-atherosclerotic activity is a modification of lipoprotein levels or LDL oxidation14. Curcumin as antioxidants could efficiently prevent LDL oxidation15. The decrease in the total cholesterol and triglycerides levels may be attributed to hypolipidemic agents in Z. montanum. The significant changes in LDL levels suggest that Z. montanum is having an effect on lipid metabolism16. It is suggested that curcumin with other chemical compounds from Z. montanum could show anti-hypercholesterolemic effects.\n\n\nConclusion\n\nIt could be concluded that Z. montanum extracts have the potenial to reduce lipid profile level, which could be further developed as a natural source of the anti-hypercholesterolemic agents.\n\n\nData availability\n\nF1000Research: Dataset 1. Effect of ethanol extract of Zingiber montanum and simvastatin in total cholesterol, triglycerides, high density lipoproteins (HDL), and low density lipoproteins (LDL) levels after 4 weeks of treatment in a high fat diet rat model., http://dx.doi.org/10.5256/f1000research.16417.d22166817", "appendix": "Grant information\n\nThis work was supported with funding from the Project Implementation Unit IsDB Universitas Mulawarman for financing this research, as part of the implementation of IsDB Grant Research Year of 2018 [448/UN17.45/DL/2018].\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nSupplementary File 1: Picture of rhizome of Zingiber montanum (J.Koenig) Link ex A.Dietr.\n\nClick here to access the data\n\n\nReferences\n\nWorld Health Organization: Raised Cholesterol: Situation and Trends. Global Health Observatory Data (GHO), 2018. Reference Source\n\nNational Institutes of Health: Hypercholesterolemia. National Library of Medicine. 2018. Reference Source\n\nLibby P: The Pathogenesis, Prevention, and Treatment of Atherosclerosis. In Harrison’s Principles of Internal Medicine, Nineteenth, DL Kasper, SL Hauser, JL Jameson, AS Fauci, DL Longo, and J Loscalzo, Eds. New York: McGraw-Hill Education, 2015; 291e1–291e10. Reference Source\n\nDraznin B, Epstein S, Turner HE, et al.: Lipids and Hyperlipidemia. In Oxford American Handbook of Endocrinology and Diabetes. New York: Oxford University Press, 2011; 677–687.\n\nIswantini D, Silitonga RF, Martatilofa E, et al.: Zingiber cassumunar, Guazuma ulmifolia, and Murraya paniculata Extracts as Antiobesity: In Vitro Inhibitory Effect on Pancreatic Lipase Activity. HAYATI J Biosci. 2011; 18(1): 6–10. Publisher Full Text\n\nThe Plant List: Zingiber montanum (J.Koenig) Link ex A.Dietr. 2018; [Online]. Reference Source\n\nKantayos V, Paisooksantivatana Y: Antioxidant Activity and Selected Chemical Components of 10 Zingiber spp. in Thailand. J Dev Sustain Agric. 2012; 7(1): 89–96. Publisher Full Text\n\nKementerian Kesehatan RI: Laporan Nasional RISTOJA 2015. Jakarta, Indonesia: Lembaga Penerbit Badan Penelitian dan Pengembangan Kesehatan, 2016. Reference Source\n\nKamazeri TS, Samah OA, Taher M, et al.: Antimicrobial activity and essential oils of Curcuma aeruginosa, Curcuma mangga, and Zingiber cassumunar from Malaysia. Asian Pac J Trop Med. 2012; 5(3): 202–209. PubMed Abstract \| Publisher Full Text\n\nBuiyan MNI, Chowdhury JU, Begum J: Volatile constituents of essential oils isolated from leaf and rhizome of Zingiber cassumunar Roxb. Bangladesh J Pharmacol. 2008; 3(2): 69–37. Publisher Full Text\n\nSingh C, Manglembi N, Swapana N, et al.: Ethnobotany, Phytochemistry and Pharmacology of Zingiber cassumunar Roxb. (Zingiberaceae). J Pharmacogn Phytochem. 2015; 4(1): 1–6. Reference Source\n\nBua-in S, Paisooksantivatana Y: Essential Oil and Antioxidant Activity of Cassumunar Ginger (Zingiberaceae: Zingiber montanum (Koenig) Link ex Dietr.) Collected from Various Parts of Thailand. Kasetsart J (Nat Sci.). 2009; 43: 467–475. Reference Source\n\nChairul C, Praptiwi P, Chairul SM: Phagocytosis Effectivity Test of Phenylbutenoid Compounds Isolated from Bangle (Zingiber cassumunar Roxb.) Rhizome. Biodiversitas. 2009; 10(1): 40–43. Publisher Full Text\n\nBei WJ, Guo J, Wu HY, et al.: Lipid-regulating effect of traditional chinese medicine: mechanisms of actions. Evid Based Complement Alternat Med. 2012; 2012: 970635. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKajal A, Kishore L, Kaur N, et al.: Therapeutic agents for the management of atherosclerosis from herbal sources. Beni-Suef Univ J Basic Appl Sci. 2016; 5(2): 156–169. Publisher Full Text\n\nArthur FKN, Woode E, Terlabi EO, et al.: Evaluation of acute and subchronic toxicity of Annona Muricata (Linn.) aqueous extract in animals. Euro J Exp Bio. 2011; 1(4): 115–124. Reference Source\n\nParamita S, Aminyoto M, Ismail S, et al.: Dataset 1 in: Anti-hypercholesterolemic effect of Zingiber montanum. extract. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16417.d221668" }	[ { "id": "40719", "date": "28 Nov 2018", "name": "Deny Susanti", "expertise": [ "Reviewer Expertise Natural Product Chemistry", "Bioactivity study" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\n1. The manuscript needs to be sent for proofreading. There are a lot of grammatical mistakes and poor sentence construction, and there is no connection between sentences in paragraphs.\n2. In the manuscript, the voucher of herbarium specimen of the plant should be stated, and also where the herbarium specimen is deposited and who is identifying the plant (botanist).\n3. Please mention the replication that was conducted in each experiment.\n4. Discussion: rewrite the discussion to discuss more details about the results of the experiments and compare the previous research or facts from the article or book, to support the results.\n\n5. Check on how to write the species name of plants for the first and subsequent usage, and be consistent.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4811", "date": "14 Aug 2019", "name": "Swandari Paramita", "role": "Author Response", "response": "1. The manuscript has sent for proofreading. 2. The plant was authenticated by Ir. Hj. Hastaniah, M.P. and the voucher specimen (voucher number: 27b/UN17.4.3.08/LL/2018) was deposited to Laboratory of Dendrology and Forest Ecology, Faculty of Forestry, Mulawarman University, Samarinda, Indonesia. 3. There was one replication that was conducted in each experiment of every sample. 4. In the discussion has been rewriting to explain the results of the experiments and compare the previous research to support the results. 5. All the species name of the plants has been corrected as “Zingiber montanum” for the first, and “Z. montanum” for subsequent usage." } ] } ]	1	https://f1000research.com/articles/7-1798
https://f1000research.com/articles/8-1433/v1	14 Aug 19	{ "type": "Systematic Review", "title": "Zika virus infection as a cause of congenital brain abnormalities and Guillain-Barré syndrome: A living systematic review", "authors": [ "Michel Jacques Counotte", "Kaspar Walter Meili", "Katayoun Taghavi", "Guilherme Calvet", "James Sejvar", "Nicola Low", "Kaspar Walter Meili", "Katayoun Taghavi", "Guilherme Calvet", "James Sejvar" ], "abstract": "Background: The Zika virus (ZIKV) caused a large outbreak in the Americas leading to the declaration of a Public Health Emergency of International Concern in February 2016. A causal relation between infection and adverse congenital outcomes such as microcephaly was declared by the World Health Organization (WHO) informed by a systematic review structured according to a framework of ten dimensions of causality, based on the work of Bradford Hill. Subsequently, the evidence has continued to accumulate, which we incorporate in regular updates of the original work, rendering it a living systematic review. Methods: We present an update of our living systematic review on the causal relation between ZIKV infection and adverse congenital outcomes and between ZIKV and GBS for four dimensions of causality: strength of association, dose-response, specificity, and consistency. We assess the evidence published between January 18, 2017 and July 1, 2019. Results: We found that the strength of association between ZIKV infection and adverse outcomes from case-control studies differs according to whether exposure to ZIKV is assessed in the mother (OR 3.8, 95% CI: 1.7-8.7, I2=19.8%) or the foetus/infant (OR 37.4, 95% CI: 11.0-127.1, I2=0%). In cohort studies, the risk of congenital abnormalities was 3.5 times higher after ZIKV infection (95% CI: 0.9-13.5, I2=0%). The strength of association between ZIKV infection and GBS was higher in studies that enrolled controls from hospital (OR: 55.8, 95% CI: 17.2-181.7, I2=0%) than in studies that enrolled controls at random from the same community or household (OR: 2.0, 95% CI: 0.8-5.4, I2=74.6%). In case-control studies, selection of controls from hospitals could have biased results. Conclusions: The conclusions that ZIKV infection causes adverse congenital outcomes and GBS are reinforced with the evidence published between January 18, 2017 and July 1, 2019.", "keywords": [ "Zika", "Disease outbreaks", "arboviruses", "causality", "Guillain-Barré syndrome", "congenital abnormalities" ], "content": "Introduction\n\nThe Zika virus (ZIKV), a mosquito-borne flavivirus, caused a large outbreak of infection in humans in the Americas between 2015–2017 (WHO Zika -Epidemiological Update). Since then, the circulation of ZIKV has decreased substantially in the Americas1 but ZIKV transmission will likely continue at a lower level2. Smaller outbreaks have been reported from countries in Africa and Asia, including Angola, India3, and Singapore4. Regions with endemic circulation, such as Thailand5, have the potential for new ZIKV outbreaks with adverse outcomes6.\n\nThe World Health Organization (WHO) declared ZIKV as a cause of adverse congenital outcomes and Guillain-Barré syndrome (GBS) as early as September 20167, informed by a systematic review of evidence structured according to a framework of ten dimensions of causality, based on Bradford Hill (Table 1)8. The accumulation of evidence on the adverse clinical outcomes of ZIKV has barely slowed down since the WHO declared the Public Health Emergency of International Concern (PHEIC) on February 1st, 2016, with approximately 250 research publications on ZIKV appearing every month (see Zika Open Access Project). We updated the systematic review to January 18, 2017 as a living systematic review by introducing automated search methods to produce a high quality, up to date, online summary of research9 about ZIKV and its clinical consequences, for all the causality dimensions10.\n\nThe latest and previous versions of this table are available as extended data16.\n\n* The causality framework is described elsewhere in detail:\n\n[https://f1000researchdata.s3.amazonaws.com/supplementary/13704/95380c74-7569-4049-bf3b-b2832794bdf9.docx].\n\nSince 2017, understanding about the pathogenesis of how ZIKV causes congenital abnormalities has evolved11,12. The quality of diagnostic methods, especially for acute ZIKV infection, has also improved13–15. More importantly, understanding of the limitations of diagnostic testing, and the need for interpretation in the context of other flavivirus infections, has developed. Important epidemiological questions about the associations between ZIKV infection and adverse congenital outcomes and GBS remain unanswered, however. Much of the early epidemiological evidence, which relied on surveillance data, was limited in use because of issues with the quality of the reporting and case definitions. The reported strength of association between ZIKV and adverse outcomes has varied in studies of different designs and in different settings. Evidence for a dose-response relationship with higher levels of exposure to ZIKV resulting in more severe outcomes, of clinical findings that are specific to ZIKV infection, or of adverse outcomes caused by different lineages of ZIKV was not found in the earlier systematic reviews.\n\nThe objective of this study is to update epidemiological evidence about associations between ZIKV infection and adverse congenital outcomes and between ZIKV and GBS for four dimensions of causality: strength of association, dose-response, specificity, and consistency.\n\n\nMethods\n\nWe performed a living systematic review, which we have described previously10. This review updates the findings of the previous reviews8,10 and will be kept up to date, in accordance with the methods described below. Reporting of the results follows the Preferred Reporting Items of Systematic reviews and Meta-Analyses (PRISMA) statement (Extended data, Supplementary File 117)18.\n\nThis review and subsequent updates will focuse on four dimensions of causality that are examined in epidemiological study designs: strength of association, dose-response relationship and specificity of effects and consistency of association (Extended data, Table 1, Supplementary File 217). Evidence for domains of causality that are typically investigated in in vitro and in vivo laboratory studies (Table 1) was not sought. In the absence of licensed vaccines or treatments for ZIKV infection, we did not search for evidence on the effects of experimental removal of ZIKV.\n\nWe considered epidemiological studies that reported original data and assessed ZIKV as the exposure and congenital abnormalities or GBS as the outcomes. We based the exposure and outcome assessment on the definitions used in the publications. We applied the following specific inclusion criteria (Extended data, Supplementary File 217):\n\nStrength of association: at the individual level, we selected studies that included participants both with and without exposure to ZIKV (Figure 1), such as cohort studies and case-control studies. At the population level, we included studies that assessed the outcome during the ZIKV outbreak and provided a comparison with pre or post-outbreak incidence of the outcome.\n\nThe latest and previous versions of this figure are available as extended data16.\n\nDose-response relationship: we included studies that assessed the relation between the level of the viral titre or the presence or severity of the symptoms and the occurrence or severity of the outcome.\n\nSpecificity of the outcome for ZIKV exposure: we included studies that assessed whether the pathological findings in cases with the outcome are specific for ZIKV infection.\n\nConsistency: we looked at eligible studies to determine the consistency of the relationship between ZIKV exposure and the outcomes across populations, study designs, regions or strains.\n\nWe searched PubMed, Embase, LILACS and databases and websites of defined health agencies (Extended data, Supplementary File 216). We included search terms for the exposure, the outcome and specific study designs. We also performed searches of the reference lists of included publications. A detailed search strategy is presented in Supplementary File 2. For this review, the search covered the period from January 19, 2017 to July 1, 2019.\n\nOne reviewer screened titles and abstracts of retrieved publications. If retained, the same reviewer screened the full text for inclusion. A second reviewer verified decisions. One reviewer extracted data from included publications into piloted extraction forms in REDCap (version 8.1.8 LTS, Research Electronic Data Capture)19. A second reviewer verified data entry. Conflicts were resolved by consulting a third reviewer.\n\nFirst, we summarised findings for each dimension of causality and for each outcome descriptively. Where available, we calculated unadjusted odds ratios (OR) or risk ratios (RR) and their 95% confidence interval (CI) from published data for unmatched study designs. For matched study designs, we used the effect measure and 95% CI presented by the authors. For publications that presented results for multiple measures of exposure and/or outcome, we compared these results. We applied the standard continuity correction of 0.5 for zero values in any cell in the two-by-two table20. We used the I² statistic to describe the percentage of variation across studies that is due to heterogeneity for reasons other than chance21. Quantitative synthesis was performed using R 3.5.122. We conducted random effects meta-analyses using the R package metafor (version 2.0-0)20. Finally, we compared descriptive and quantitative findings from this review period with previous versions of the review8,10.\n\nDaily searches of PubMed, Embase and LILACS are automated and monthly searches are performed manually for other information sources in the first week of the month (Extended data, Supplementary File 217), with screening of all retrieved publications on the same day. The search strategy consisted of a combination of free terms and MESH terms that identified the exposure and outcomes (Extended data, Supplementary File 217). Searches from multiple sources were combined and automatically deduplicated by an algorithm that was tested against manual deduplication. Unique records enter a central database, and reviewers are notified of new content.\n\nThe tables and figures presented in this paper will be updated every six months as a new version of this publication. As soon as new studies are included, their basic study characteristics are extracted and provided online https://zika.ispm.unibe.ch/apps/causalitymap/.\n\nWe will keep the living systematic review up to date for as long as new relevant data are published and at least until October 31, 2021, the end date of the project funding.\n\nTo assess the risk of bias of cohort studies and case-control studies, we compiled a list of questions in the domains of selection bias, information bias, and confounding, based on the quality appraisal checklist of the United Kingdom National Institute for Health and Care Excellence (NICE) and literature23. Two independent reviewers conducted the quality assessment. Disagreements were resolved by a third reviewer.\n\n\nResults\n\nFrom January 19, 2017 to July 1, 2019 we screened 1941 publications, of which we included 638 based on title and abstract. After reviewing the full text, 249 publications were included (Table 2, Figure 3). Of these publications, 195 reported on congenital abnormalities linked to ZIKV24–217 and 59 on GBS4,44,118,201,203,206,218–270. Five outbreak reports described both outcomes44,118,201,203,206.\n\nThe latest and previous versions of this table are available as extended data16.\n\n* Baseline review, earliest date of each information source to May 30, 20168;\n\n† Update 1, May 30, 2016 to January 18, 201710.\n\nThe size of the points correspond with the number of exposed individuals with the adverse outcome, according to the definitions used in the publications. The latest and previous versions of this figure are available as extended data16.\n\nAdapted from: Moher et al. (2009)18. The latest and previous versions of this figure are available as extended data16.\n\nWe included 39 case reports24–61, 62 case series62–123, 10 case-control studies124–133, 35 cohort studies134–168, 19 cross-sectional studies169–187, seven ecological studies188–194, three modelling studies195–197 and 20 outbreak reports198–217 that report on congenital abnormalities linked to ZIKV.\n\nCausality dimensions\n\nStrength of association. Individual level: In this review period, five case-control studies reported on strength of association, four in Brazil (n=670 participants)125,126,128,130 and one in French Polynesia (n=123 participants)131. The studies assess adverse pregnancy outcomes including infants born with microcephaly, according to exposure to ZIKV for cases. Of these, all studies matched controls, based on gestational age and/or region. During the review period up to January 18, 2017, we included one case-control study271, which we replaced with a publication reporting the final results of the study126. The meta-analyses incorporate estimates from studies identified in all review periods.\n\nAssessment of exposure status varied between the studies (Extended data, Supplementary File 317). In five case-control studies, exposure to ZIKV was assessed in the mother, based on clinical symptoms of ‘suspected Zika virus infection’125, or presence of maternal antibodies measured by IgM (Kumar et al. (2016)272), PRNT (de Araujo et al. (2018)126, Subissi et al. (2018)131), or both PRNT and IgG (Moreira-Soto et al. (2018)) maternal antibody127.\n\nIn meta-analysis, we found that the odds of adverse congenital outcomes (microcephaly or congenital abnormalities) were 3.8 times higher in ZIKV-infected mothers (95% CI: 1.7-8.7, tau2=0.18, I2=19.8%, Figure 4). Moreira-Soto et al. (2018) found that in Bahia, Brazil, Chikungunya infection was also associated with being a case127.\n\nForest plot and meta-analysis of case-control studies reporting on ZIKV infection assessed in mothers (A) and in infants (B) and adverse congenital outcomes (microcephaly, congenital malformations, central nervous system abnormalities). The odds ratio from the five case-control studies that assess exposure in mothers combined is 3.8 (95% CI: 1.7-8.7, tau2=2.37, I2=19.8%); the odds ratio for the studies that assess exposure in infants is 37.4 (95% CI: 11.0-127.1, tau2=0, I2=0%). The odds ratios are plotted on the log scale. Abbreviations: CSF, cerebrospinal fluid, PRNT, plaque reduction neutralisation test; RE, random effects; RT-PCR, reverse transcription polymerase chain reaction. The latest and previous versions of this figure are available as extended data16.\n\nIn two matched case-control studies, exposure to ZIKV was assessed in infants; Araujo et al. found a 73.1 (95% CI 13·0–Inf) times higher odds was reported for microcephaly when ZIKV infection was assessed by reverse transcription polymerase chain reaction (RT-PCR) in the neonate126. Krow-Lucal et al. (2018) found an OR of 21.9 (95% CI: 7.0-109.3) based on evidence of recent Zika infection assessed using IgM followed by PRNT in infants in Paraiba, Brazil128. When exposure was assessed at the infant-level, the combined odds of adverse congenital outcomes was 37.4 times higher (95% CI: 11.0-127.1, tau2=0, I2=0%, Figure 4).\n\nIn this review period, one cohort study reported on strength of association, in 610 pregnant women returning from ZIKV-affected areas in Central and South America to the USA138. Maternal ZIKV exposure was measured using RT-PCR or IgM followed by plaque reduction neutralisation test (PRNT). Among the 28 infants born to ZIKV-infected mothers, none was diagnosed with microcephaly and, one was born with a major malformation. In the ZIKV-unexposed group, eight out of 306 had major malformations. A complete overview of different outcomes assessed is presented in the extended data, Supplementary File 317. During the review period up to January 18, 2017, we included two cohort studies, one in women with rash and fever (Brasil et al. (2016)) and one in unselected pregnant women (Pomar et al. (2017))273,274. In meta-analysis of all three studies, we found that the risk of microcephaly was 3.5 times higher in ZIKV-infected mothers of babies (95% CI: 0.90-13.51, tau2=0, I2=0%, Figure 5).\n\nThe risk ratio from the random effects model is 3.5 (95% CI: 0.9-13.5, tau2=0,I2=0%). The risk ratios are provided on the log scale. Abbreviations: ZIKV, Zika virus; PRNT, plaque reduction neutralisation test; RE, random effects; RT-PCR, Reverse transcription polymerase chain reaction. The latest and previous versions of this figure are available as extended data.\n\nPopulation level: At a population level, data from Mexico collected at different altitudes during the ZIKV outbreak, showed that the risk of microcephaly was increased in regions at altitudes below 2200m, in which ZIKV can circulate196. Hay et al. (2018) reanalysed surveillance data from Colombia and northeast Brazil and concluded that time-dependent reporting changes might have caused apparent inconsistencies in the proportion of congenital abnormalities as a result of maternal ZIKV infection197.\n\nDose response. Halai et al. (2017)120 examined the severity of congenital outcomes according to measures of the severity of maternal ZIKV infection in a subset of mothers in the cohort presented by Brasil et al. (2016)274. They evaluated ZIKV load, assessed by RT-PCR using the cycle threshold (CT) as a measure of number of RNA copies, and a severity score of symptoms in 131 pregnant women. They concluded that neither higher viral load nor more severe symptoms was associated with more severe congenital abnormalities136. Moreira-Soto et al. found higher maternal antibody titers in microcephaly cases compared with controls127. In previous review periods, Honein et al. (2016) compared outcomes in neonates born to symptomatic and asymptomatic infected pregnant women returning to the USA with possible ZIKV infection and found no differences275.\n\nSpecificity. Although some outcomes, such as lingual phenotype177 or neurogenic bladder276, have been hypothesised as a specific phenotype for congenital ZIKV infection, no additional evidence was identified that certain congenital adverse findings are specific for congenital ZIKV infection.\n\nConsistency. Geographical region: All four WHO geographic regions (the Africa region [AFRO], the American region [AMRO], the South-East Asian region [SEARO] and the Western Pacific region [WPRO]) with past or active ZIKV transmission have now reported congenital abnormalities due to ZIKV infection. During this review period, the first congenital abnormality due to infection with the Asian lineage of the virus on the African mainland occurred in a traveller returning from Angola47. Possible cases of congenital abnormalities have occurred in Guinea-Bissau96. In the most recent WHO situation report from March 2017, two cases of microcephaly are documented in Thailand and one in Vietnam, which were also described in detail in other works24,54,107. We identified another publication on congenital abnormalities due to endemic ZIKV in Cambodia110. The occurrence of congenital adverse outcomes in AFRO, SEARO, and WPRO seems sporadic, despite the endemic circulation of ZIKV. As noted above, the observed complication rate varied strongly between regions. Extended data, Supplementary File 3 provides a full overview of the published studies on congenital abnormalities per region and country17.\n\nTraveller/non-traveller populations: In this update, we found further evidence that congenital abnormalities occurred in infants born to women travellers returning from ZIKV-affected areas and women remaining in those areas. In total, 25 publications report on 272 congenital abnormalities due to ZIKV infection in travellers27,29,30,32,34,36,38,42,45,56,58,61,89,94,98,109,117,122,123,137,140,151,153,166,173, with 109 publications reporting congenital abnormalities due to ZIKV in 2652 non travellers24–26,28,31,33,35,37,39–41,43,44,46–48,50,51,54,57,60,62–79,81,83–85,87,88,90–93,95–97,100,101,104–106,110–113,115,116,118,119,121,124–135,142–147,149,152,154,156–164,167,168,171,172,175–180,182,186\n\nIn this review period, evidence emerged that transmission through sexual contact with infected travellers also resulted in foetal infection58,59.\n\nStudy designs: The association between ZIKV infection and congenital abnormalities was consistent across different study designs (Table 2).\n\nLineages: We found no new evidence of consistency across different lineages from observational studies. The currently observed adverse congenital outcomes are linked to the ZIKV of the Asian lineage.\n\nIn all case-control studies, uncertainty about the exposure status due to imperfect tests could result in a bias towards the null. Some studies might suffer from recall bias where exposure was assessed by retrospectively asking about symptoms125,131. For the cohort studies138,274, the enrolment criteria were based on symptomatology. As a result, even in the absence of evidence of ZIKV, the unexposed groups might have had conditions that were unfavourable to their pregnancy. We expect this to bias the results towards the null or underestimate the true effect. Owing to imperfect diagnostic techniques, both false positives (IgM, cross reactivity) and false negatives (due to the limited detection window for RT-PCR) might occur, potentially resulting in bias; the direction of this bias would often be towards the null. None of the studies controlled for potential confounding. Extended data, Supplementary File 4 provides the full risk of bias assessment of the studies included in the meta-analysis17.\n\nDuring this review period, we included 17 case reports44,218–233, 22 case series118,234–254, seven case-control studies4,255–260, one ecological study264, one modelling study265 and eight outbreak reports201,203,206,266–270 that reported on ZIKV infection and GBS.\n\nDuring this review period, we included 17 case reports46,220–235, 22 case series120,236–256, seven case-control studies4,257–262, one ecological study266, one modelling study265 and eight outbreak reports201,203,206,266–270 that reported on ZIKV infection and GBS.\n\nCausality dimensions\n\nStrength of association. Individual level. The number of studies reporting on the strength of association between ZIKV infection and GBS at an individual level increased substantially. We identified five case-control studies255–259 published since the previous update, which included one case-control study from French Polynesia277. All studies were matched for age and place of residence. In the studies from Brazil, Colombia, Puerto Rico and New Caledonia, temporal clustering of cases in association with ZIKV circulation was documented255–258. In Bangladesh, ZIKV transmission was endemic259. Exposure assessment was based on serology255,256 or a combination of RT-PCR and serology257–259. Extended data, Supplementary File 3 shows the variability in ORs according to criteria for ZIKV exposure assessment, based on unmatched crude data extracted from each case-control study17.\n\nFigure 6 shows the association between GBS and ZIKV infection, using the diagnostic criteria that were most similar across studies. Heterogeneity was considerable (I2=78.3%), but was reduced slightly after stratification based on the method of selection of controls. The summary OR was higher in studies that enrolled controls from hospital (OR: 55.8, 95% CI: 17.2-181.7, tau2=0, I2=0%)258,277 than in studies that enrolled controls at random from within the same community255–257 or from the same household259 (OR: 2.0, 95% CI: 0.8-5.4, tau2=0.46, I2=74.6%). Amongst studies with community controls, ORs were lower when enrolment and assessment took place several months after onset of symptoms255,256 than in studies with contemporaneous enrolment257,259. To further illustrate the heterogeneity in exposure assessment between and within the studies, we provide additional aggregations of the data in Extended data, Supplementary File 317.\n\nOdds ratios (ORs) are shown on the log-scale. The meta-analysis is stratified by the selection of controls: Hospital controls, or community/household controls. Most similar exposure assessment measures are compared (IgM256,258,259,277, recent flavivirus infection255, or IgM and/or RT-PCR257). OR: 7.0 [95% CI: 1.7-28.8, tau2=2.78, I2=78.3%]. ORs from studies marked with an asterisk (*) are matched ORs, unmarked studies provided crude ORs. The latest and previous versions of this figure are available as extended data16.\n\nPopulation level. At a population level, Mier-Y-Teran-Romero et al. (2018) found that the estimated incidence of GBS ranged between 1.4 (0.4–2.5) and 2.2 (0.8–5.0) per 10,000 ZIKV infections comparing surveillance/reported cases from Brazil, Colombia, Dominican Republic, El Salvador, French, Honduras, Puerto Rico, Suriname, Venezuela, and Micronesia. The across-location minimum and maximum estimates were used to estimate an average risk of having GBS and being reported after ZIKV infection across locations of approximately 2.0 GBS cases per 10,000 infections (95% credible interval 0.5–4.5 per 10,000 ZIKV infections)265.\n\nDose response. In a case-control study, Lynch et al. (2018) found higher titres of neutralising antibodies in ZIKV-infected GBS cases than in patients with symptomatic ZIKV infection but without GBS260.\n\nSpecificity. Dirlikov et al. (2018) compared Puerto Rican GBS cases reported through public health surveillance that were preceded by ZIKV and cases that were not preceded by ZIKV infection249. Clinical features involving cranial nerves were observed more frequently in ZIKV-related cases and, at a six-month follow-up visit, residual cranial neuropathy was noted more often in this group. However, clinical symptoms did not allow a distinction to be made between ZIKV and non-ZIKV related GBS.\n\nConsistency. Geographical region: During this review period, GBS likely due to ZIKV infection was reported in Asia; including Thailand, Bangladesh, Singapore and India4,48,252,259. Publication in the WHO Region of the Americas followed the pattern as observed before and no GBS linked to ZIKV infection was reported in Africa. Extended data, Supplementary File 3 provides a full overview of the published studies on congenital abnormalities by region and country17. In a reanalysis of surveillance data from the Region of the Americas, Ikejezie et al. (2016) found consistent time trends between GBS incidence and ZIKV incidence264.\n\nTraveller/non-traveller populations: In studies included in this update, we found additional evidence of GBS in both travellers and non-travellers with ZIKV infection. Ten publications report on 11 travellers218,220,222–224,227,229,232–234, while 34 publications report GBS due to ZIKV in 402 non travellers4,44,118,219,221,225,226,228,230,231,235–237,239–247,249,251,253–260,262,263.\n\nStudy designs: Across the different study designs, the relation between GBS and ZIKV is consistently shown. Table 2 and Extended data, Supplementary File 3 provide an overview of the included study designs17.\n\nLineages: We still lack evidence on the consistency of the relation between GBS and ZIKV across different lineages from observational studies. The observed cases of GBS were linked to ZIKV of the Asian lineage.\n\nPotential selection bias in case-control studies was introduced by the selection of controls from hospitals rather than from the communities in which the cases arose258,277. Uncertainty about the exposure status due to imperfect tests would tend to result in a bias towards the null. Two case-control studies did not conduct a matched analysis although controls were matched, and no study controlled for potential confounding by factors other than those used for matching. Exclusion criteria and participation rate, especially of the controls, were poorly reported. Extended data, Supplementary File 4 provides the full risk of bias assessment of the studies included in the meta-analysis17.\n\n\nDiscussion\n\nIn this living systematic review, we summarised the evidence from 249 observational studies in humans on four dimensions of the causal relationship between ZIKV infection and adverse congenital outcomes and GBS, published between January 18, 2017 and July 1, 2019.\n\nThe strengths of this living systematic review are, first, that we automated much of the workflow10; we searched both international and regional databases daily and we screen papers for eligibility as they became available, so publication bias is unlikely. Second, we have quantified the strength of association between ZIKV infection and congenital abnormalities and GBS and investigated heterogeneity of outcome and exposure assessment within and between studies. Third, for congenital outcomes, we included studies with both microcephaly and other possible adverse outcomes, acknowledging the spectrum of congenital adverse outcomes caused by ZIKV. This work also has several limitations. First, we have not assessed the dimensions of the causality framework that involve laboratory studies, so we have not updated the pathobiology of ZIKV complications, which was addressed in the baseline review8 and the first update to January 201710. Limiting the review to epidemiological domains has allowed more detailed analyses of these studies and we hope that laboratory scientists will continue to review advances in these domains. Second, the rate of publications on ZIKV remains high so, despite the reduced scope and automation, maintenance of the review is time-consuming and data extraction cannot be automated. Third, this review may suffer from continuity bias, which is important for the conduct and interpretation of living systematic reviews and results from changes in the author team. Careful adherence to the protocol will reduce this risk.\n\nZIKV and congenital abnormalities: Since the earlier versions of the review8,10, evidence on the causal relationship between ZIKV infection and congenital abnormalities has expanded. Unfortunately, the total number of cases investigated in the published cohort or case-control studies remains small. In case-control studies in which infants with microcephaly or other congenital abnormalities are compared with unaffected infants, the strength of association differs according to whether exposure to ZIKV is assessed in in the mother (OR 3.8, 95% CI: 1.7-8.7, tau2=0.18, I2=19.8%) or the foetus/infant (OR 37.4, 95% CI: 11.0-127.1, tau2=0, I2=0%). This large difference in effect size can be attributed to the fact that not all maternal ZIKV infections result in foetal infection. In cohort studies, the risk of congenital abnormalities was 3.5 times higher (95% CI: 0.9-13.5, I2=0%, tau2=0) in mothers with evidence of ZIKV infection than without, which is similar to the OR for maternal exposure to ZIKV estimated from case-control studies. Further research is needed to understand the drivers of mother to child transmission. Higher maternal antibody titres were correlated with a higher incidence of adverse congenital outcomes in one case-control study127. However, amongst ZIKV-infected mothers followed prospectively, severity of ZIKV infection was not associated with more severe congenital abnormalities136. Convincing evidence on a dose-response relation is therefore still lacking.\n\nZIKV and GBS: Evidence on the causal relation between ZIKV infection and GBS has grown since our last review10. The body of evidence is still smaller than that for congenital abnormalities, possibly because GBS is a rare complication, estimated to occur in 0.24 per 1000 ZIKV infections277. In this review, the strength of association between GBS and ZIKV infection, estimated in case-control studies, tended to be lower than observed in the first case-control study reported by Cao-Lormeau (2016) in French Polynesia277. It is possible that the finding by Cao-Lormeau et al. was a ‘random high’, a chance finding278. Simon et al., however, found a similarly strong association in a case-control study in New Caledonia258. In both these studies, controls were patients in the same hospital. Although matched for place of residence, it is possible that they were less likely to have been exposed to ZIKV than the cases, resulting in an overestimation of the OR. In case-control studies in which controls were enrolled from the same communities as the cases, estimated ORs were lower, presumably because exposure to ZIKV amongst community-enrolled controls is less biased than amongst hospital controls279. Under-ascertainment of ZIKV infection in case-control studies in which enrolment occurred several months after the onset of symptoms255,256 is also likely to have reduced the observed strength of association. There is also possible evidence of a dose-response relationship, with higher levels of neutralising antibodies to both ZIKV and dengue in people with GBS260. However, the level of antibody titre might not be an appropriate measure of viral titre, and merely a reflection of the intensity of the immune response. Taking into account the entire body of evidence, inference to the best explanation280 supports the conclusion that ZIKV is a cause of GBS. The prospect of more precise and robust estimates of the strength of association between ZIKV and GBS is low because outbreaks need to be sufficiently large to enrol enough people with GBS. In the large populations that were exposed during the 2015–2017 outbreak, herd immunity will limit future ZIKV outbreaks.\n\nThe sample sizes of studies published to date are smaller than those recommended by WHO for obtaining precise estimates of associations between ZIKV and adverse outcomes [Harmonization of ZIKV Research Protocols to Address Key Public Health]. Given the absence of large new outbreaks of ZIKV infection in 2017–2019, there is a need for consortia of researchers to analyse their data in meta-analyses based on individual participant data [Individual Participant Data Meta-analysis of Zika-virus related cohorts of pregnant women (ZIKV IPD-MA)]. Future collaborative efforts will help to quantify the absolute risks of different adverse congenital outcomes and allow investigation of heterogeneity between studies136,149,275.\n\nThis review highlights additional research gaps. We did not assess the complication rates within the infected group in studies without an unexposed comparison group; the adverse outcomes are not pathognomonic for ZIKV infection, making an appropriate comparison group necessary. Although there are no individual features of ZIKV infection that are completely specific, the growing number of publications on ZIKV will allow better ascertainment of the features of a congenital Zika syndrome281. In this review, we did not take into account the performance of the diagnostic tests in assessing the strength of association. Future research should include robust validation studies, and improved understanding of contextual factors in the performance of diagnostic tests, including the influence of previous circulation of other flaviviruses, the prevalence of ZIKV and the test used.\n\nThis living systematic review will continue to follow studies of adverse outcomes originating from ZIKV circulation in the Americas, but research in regions with endemic circulation of ZIKV is expected to increase. Such studies will clarify whether ZIKV circulation in Africa and Asia also results in adverse outcomes, as suggested by the case-control study of GBS from Bangladesh259. Increased awareness might improve the evidence-base in these regions, where misperceptions about the potential risks of ZIKV-associated disease with different virus lineages has been reported282. An important outstanding question remains whether the absence of reported cases of congenital abnormalities or GBS in these regions represent a true absence of complications or is this due to weaker surveillance systems or reporting283. The conclusions that ZIKV infection causes adverse congenital outcomes and GBS are reinforced with the evidence published between January 18, 2017 and July 1, 2019.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\nHarvard dataverse: Living systematic review on adverse outcomes of Zika - Supplementary Material. https://doi.org/10.7910/DVN/S7USUI17\n\nThis project contains the following extended data:\n\nSupplementaryFile1Prisma.docx (PRISMA checklist)\n\nSupplementaryFile2Methods.docx (Supplementary file 2, additional information to the Methods)\n\nSupplementaryFile3Results.docx (Supplementary file 3, additional information to the Results)\n\nSupplementaryFile4ROB.tab (Risk of bias assessment)\n\nHarvard dataverse: Living systematic review on adverse outcomes of Zika - Figures and Table. https://doi.org/10.7910/DVN/DLP5AN16\n\nThis project contains the following extended data:\n\nFig1.pdf (Most recent version of Figure 1)\n\nFig2.pdf (Most recent version of Figure 3, PRISMA flowchart)\n\nFig3A.pdf (Most recent version of Figure 4A)\n\nFig3B.pdf (Most recent version of Figure 4B)\n\nFig4.pdf (Most recent version of Figure 5)\n\nFig5.pdf (Most recent version of Figure 6)\n\nTable1.pdf (Most recent version of Table 1)\n\nTable2.pdf (Most recent version of Table 2)\n\nPRISMA checklist and flow diagram for ‘Zika virus infection as a cause of congenital brain abnormalities and Guillain-Barré syndrome: A living systematic review’, https://doi.org/10.7910/DVN/S7USUI and Figure 2.", "appendix": "Grant information\n\nThis work was supported by the Swiss National Science Foundation [320030_170069], end date: 31/10/2021.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nDisclaimer\n\nThe views expressed in this article are those of the authors and do not necessarily represent the official positions of the Centers for Disease Control and Prevention.\n\n\nReferences\n\nButler D: Drop in cases of Zika threatens large-scale trials. Nature. 2017; 545(7655): 396–7. PubMed Abstract \| Publisher Full Text\n\nColón-González FJ, Peres CA, Steiner São Bernardo C, et al.: After the epidemic: Zika virus projections for Latin America and the Caribbean. PLoS Negl Trop Dis. 2017; 11(11): e0006007. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHamer DH, Chen LH: Zika in Angola and India. J Travel Med. 2019; 26(5): taz012. PubMed Abstract \| Publisher Full Text\n\nUmapathi T, Kam YW, Ohnmar O, et al.: The 2016 Singapore Zika virus outbreak did not cause a surge in Guillain-Barré syndrome. J Peripher Nerv Syst. 2018; 23(3): 197–201. PubMed Abstract \| Publisher Full Text\n\nRuchusatsawat K, Wongjaroen P, Posanacharoen A, et al.: Long-term circulation of Zika virus in Thailand: an observational study. Lancet Infect Dis. 2019; 19(4): 439–46. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAlthouse BM, Vasilakis N, Sall AA, et al.: Potential for Zika Virus to Establish a Sylvatic Transmission Cycle in the Americas. PLoS Negl Trop Dis. 2016; 10(12): e0005055. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHeymann DL, Hodgson A, Sall AA, et al.: Zika virus and microcephaly: why is this situation a PHEIC? Lancet. 2016; 387(10020): 719–21. PubMed Abstract \| Publisher Full Text\n\nKrauer F, Riesen M, Reveiz L, et al.: Zika Virus Infection as a Cause of Congenital Brain Abnormalities and Guillain-Barré Syndrome: Systematic Review. PLoS Med. 2017; 14(1): e1002203. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nElliott JH, Turner T, Clavisi O, et al.: Living systematic reviews: an emerging opportunity to narrow the evidence-practice gap. PLoS Med. 2014; 11(2): e1001603. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCounotte M, Egli-Gany D, Riesen M, et al.: Zika virus infection as a cause of congenital brain abnormalities and Guillain-Barré syndrome: From systematic review to living systematic review [version 1; peer review: 2 approved, 1 approved with reservations]. F1000Res. 2018; 7: 196. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDevakumar D, Bamford A, Ferreira MU, et al.: Infectious causes of microcephaly: epidemiology, pathogenesis, diagnosis, and management. Lancet Infect Dis. 2018; 18(1): e1–e13. PubMed Abstract \| Publisher Full Text\n\nFrenkel LD, Gomez F, Sabahi F: The pathogenesis of microcephaly resulting from congenital infections: why is my baby's head so small? Eur J Clin Microbiol Infect Dis. 2018; 37(2): 209–26. PubMed Abstract \| Publisher Full Text\n\nKoopmans M, de Lamballerie X, Jaenisch T, et al.: Familiar barriers still unresolved-a perspective on the Zika virus outbreak research response. Lancet Infect Dis. 2019; 19(2): e59–e62. PubMed Abstract \| Publisher Full Text\n\nFischer C, Pedroso C, Mendrone A Jr, et al.: External Quality Assessment for Zika Virus Molecular Diagnostic Testing, Brazil. Emerg Infect Dis. 2018; 24(5): 888–892. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCharrel R, Mogling R, Pas S, et al.: Variable Sensitivity in Molecular Detection of Zika Virus in European Expert Laboratories: External Quality Assessment, November 2016. J Clin Microbiol. 2017; 55(11): 3219–26. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCounotte MJ: Living systematic review on adverse outcomes of Zika - Figures and Table. Harvard Dataverse, V3. 2019. http://www.doi.org/10.7910/DVN/DLP5AN\n\nCounotte MJ: Living systematic review on adverse outcomes of Zika - Supplementary Material. Harvard Dataverse, V1, UNF:6:sAGbGceAYGoHLIT7sDtDsw== [fileUNF]. 2019. http://www.doi.org/10.7910/DVN/S7USUI\n\nMoher D, Liberati A, Tetzlaff J, et al.: Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009; 6(7): e1000097. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHarris PA, Taylor R, Thielke R, et al.: Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009; 42(2): 377–81. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nViechtbauer W: Conducting Meta-Analyses in R with the metafor Package. J Stat Softw. 2010; 36(3): 1–48. Publisher Full Text\n\nHiggins JP, Thompson SG: Quantifying heterogeneity in a meta-analysis. Stat Med. 2002; 21(11): 1539–58. PubMed Abstract \| Publisher Full Text\n\nR Core Team: R: A Language and Environment for Statistical Computing. Vienna, Austria. 2018. Reference Source\n\nHernán MA, Hernández-Díaz S, Robins JM: A structural approach to selection bias. Epidemiology. 2004; 15(5): 615–25. PubMed Abstract \| Publisher Full Text\n\nMoi ML, Nguyen TT, Nguyen CT, et al.: Zika virus infection and microcephaly in Vietnam. Lancet Infect Dis. 2017; 17(8): 805–6. PubMed Abstract \| Publisher Full Text\n\nSantos VS, Oliveira SJG, Gurgel RQ, et al.: Case Report: Microcephaly in Twins due to the Zika Virus. Am J Trop Med Hyg. 2017; 97(1): 151–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMattar S, Ojeda C, Arboleda J, et al.: Case report: microcephaly associated with Zika virus infection, Colombia. BMC Infect Dis. 2017; 17(1): 423. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZacharias N, Whitty J, Noblin S, et al.: First Neonatal Demise with Travel-Associated Zika Virus Infection in the United States of America. AJP Rep. 2017; 7(2): e68–e73. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBenjamin I, Fernández G, Figueira JV, et al.: Zika virus detected in amniotic fluid and umbilical cord blood in an in vitro fertilization-conceived pregnancy in Venezuela. Fertil Steril. 2017; 107(6): 1319–22. PubMed Abstract \| Publisher Full Text\n\nKaur G, Manganas L: EEG findings in a case of congenital zika virus syndrome. Neurology. 2017; 88(16 Supplement 1). Reference Source\n\nSaulino D, Gaston E, Younke B, et al.: The first zika-related infant mortality in the United States: An autopsy case report. Laboratory Investigation. 2017; 97: 10A.\n\nZuanazzi D, Arts EJ, Jorge PK, et al.: Postnatal Identification of Zika Virus Peptides from Saliva. J Dent Res. 2017; 96(10): 1078–84. PubMed Abstract \| Publisher Full Text\n\nLovagnini Frutos MG, Ochoa JH, Barbás MG, et al.: New Insights into the Natural History of Congenital Zika Virus Syndrome. Fetal Diagn Ther. 2018; 44(1): 72–6. PubMed Abstract \| Publisher Full Text\n\nVillamil-Gomez WE, Guijarro E, Castellanos J, et al.: Congenital Zika syndrome with prolonged detection of Zika virus RNA. J Clin Virol. 2017; 95: 52–4. PubMed Abstract \| Publisher Full Text\n\nRodó C, Suy A, Sulleiro E, et al.: In utero negativization of Zika virus in a foetus with serious central nervous system abnormalities. Clin Microbiol Infect. 2018; 24(5): 549 e1–e3. PubMed Abstract \| Publisher Full Text\n\nJucá E, Pessoa A, Ribeiro E, et al.: Hydrocephalus associated to congenital Zika syndrome: does shunting improve clinical features? Childs Nerv Syst. 2018; 34(1): 101–6. PubMed Abstract \| Publisher Full Text\n\nRaymond A, Jakus J: Cerebral Infarction and Refractory Seizures in a Neonate with Suspected Zika Virus Infection. Pediatr Infect Dis J. 2018; 37(4): e112–e4. PubMed Abstract\n\nRabelo K, de Souza Campos Fernandes RC, de Souza LJ, et al.: Placental Histopathology and Clinical Presentation of Severe Congenital Zika Syndrome in a Human Immunodeficiency Virus-Exposed Uninfected Infant. Front Immunol. 2017; 8: 1704. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAngelidou A, Michael Z, Hotz A, et al.: Is There More to Zika? Complex Cardiac Disease in a Case of Congenital Zika Syndrome. Neonatology. 2018; 113(2): 177–82. PubMed Abstract \| Publisher Full Text\n\nde Freitas Ribeiro BN, Muniz BC, Gasparetto EL, et al.: Congenital involvement of the central nervous system by the Zika virus in a child without microcephaly - spectrum of congenital syndrome by the Zika virus. J Neuroradiol. 2018; 45(2): 152–3. PubMed Abstract \| Publisher Full Text\n\nChimelli L, Moura Pone S, Avvad-Portari E, et al.: Persistence of Zika Virus After Birth: Clinical, Virological, Neuroimaging, and Neuropathological Documentation in a 5-Month Infant With Congenital Zika Syndrome. J Neuropathol Exp Neurol. 2018; 77(3): 193–8. PubMed Abstract \| Publisher Full Text\n\nRegadas VC, Silva MCE, Abud LG, et al.: Microcephaly caused by congenital Zika virus infection and viral detection in maternal urine during pregnancy. Rev Assoc Med Bras (1992). 2018; 64(1): 11–4. PubMed Abstract \| Publisher Full Text\n\nSchwartz KL, Chan T, Rai N, et al.: Zika virus infection in a pregnant Canadian traveler with congenital fetal malformations noted by ultrasonography at 14-weeks gestation. Trop Dis Travel Med Vaccines. 2018; 4: 2. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPrata-Barbosa A, Cleto-Yamane TL, Robaina JR, et al.: Co-infection with Zika and Chikungunya viruses associated with fetal death-A case report. Int J Infect Dis. 2018; 72: 25–7. PubMed Abstract \| Publisher Full Text\n\nRabelo K, Souza LJ, Salomão NG, et al.: Placental Inflammation and Fetal Injury in a Rare Zika Case Associated With Guillain-Barré Syndrome and Abortion. Front Microbiol. 2018; 9: 1018. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGuevara JG, Agarwal-Sinha S: Ocular abnormalities in congenital Zika syndrome: a case report, and review of the literature. J Med Case Rep. 2018; 12(1): 161. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGiovanetti M, Goes de Jesus J, Lima de Maia M, et al.: Genetic evidence of Zika virus in mother's breast milk and body fluids of a newborn with severe congenital defects. Clin Microbiol Infect. 2018; 24(10): 1111–2. PubMed Abstract \| Publisher Full Text\n\nSassetti M, Zé-Zé L, Franco J, et al.: First case of confirmed congenital Zika syndrome in continental Africa. Trans R Soc Trop Med Hyg. 2018; 112(10): 458–62. PubMed Abstract \| Publisher Full Text\n\nBrito CAA, Henriques-Souza A, Soares CRP, et al.: Persistent detection of Zika virus RNA from an infant with severe microcephaly - a case report. BMC Infect Dis. 2018; 18(1): 388. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGunturiz ML, Cortés L, Cuevas EL, et al.: Congenital cerebral toxoplasmosis, Zika and chikungunya virus infections: a case report. Biomedica. 2018; 38(2): 144–52. PubMed Abstract \| Publisher Full Text\n\nValdespino-Vázquez MY, Sevilla-Reyes EE, Lira R, et al.: Congenital Zika Syndrome and Extra-Central Nervous System Detection of Zika Virus in a Pre-term Newborn in Mexico. Clin Infect Dis. 2019; 68(6): 903–12. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVentura CV, Bandstra ES, Fernandez MP, et al.: First Locally Acquired Congenital Zika Syndrome Case in the United States: Neonatal Clinical Manifestations. Ophthalmic Surg Lasers Imaging Retina. 2018; 49(9): e93–e8. PubMed Abstract \| Publisher Full Text\n\nLemos de Carvalho A, Brites C, Taguchi TB, et al.: Congenital Zika Virus Infection with Normal Neurodevelopmental Outcome, Brazil. Emerg Infect Dis. 2018; 24(11): 2128–30. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHo CY, Castillo N, Encinales L, et al.: Second-trimester Ultrasound and Neuropathologic Findings in Congenital Zika Virus Infection. Pediatr Infect Dis J. 2018; 37(12): 1290–3. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWongsurawat T, Jenjaroenpun P, Athipanyasilp N, et al.: Genome Sequences of Zika Virus Strains Recovered from Amniotic Fluid, Placenta, and Fetal Brain of a Microcephaly Patient in Thailand, 2017. Microbiol Resour Announc. 2018; 7(11): pii: e01020-18. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRibeiro BNF, Marchiori E: Congenital Zika syndrome associated with findings of cerebellar cortical dysplasia - Broadening the spectrum of presentation of the syndrome. J Neuroradiol. 2018; pii: S0150-9861(18)30261-X. PubMed Abstract \| Publisher Full Text\n\nLockrow J, Tully H, Saneto RP: Epileptic spasms as the presenting seizure type in a patient with a new \"O\" of TORCH, congenital Zika virus infection. Epilepsy Behav Case Rep. 2018; 11: 1–3. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDavila-Castrodad NM, Reyes-Bou Z, Correa-Rivas M, et al.: First Autopsy of a Newborn with Congenital Zika Syndrome in Puerto Rico. P R Health Sci J. 2018; 37(Special Issue): S81–S4. PubMed Abstract\n\nYarrington CD, Hamer DH, Kuohung W, et al.: Congenital Zika syndrome arising from sexual transmission of Zika virus, a case report. Fertil Res Pract. 2019; 5: 1. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKhatib A, Showler AJ, Kain D, et al.: A diagnostic gap illuminated by a sexually-transmitted case of congenital Zika virus infection. Travel Med Infect Dis. 2019; 27: 117–8. PubMed Abstract \| Publisher Full Text\n\nSantos GR, Pinto CAL, Prudente RCS, et al.: Case Report: Histopathologic Changes in Placental Tissue Associated With Vertical Transmission of Zika Virus. Int J Gynecol Pathol. 2019. PubMed Abstract \| Publisher Full Text\n\nSulleiro E, Frick MA, Rodó C, et al.: The challenge of the laboratory diagnosis in a confirmed congenital Zika virus syndrome in utero: A case report. Medicine (Baltimore). 2019; 98(20): e15532. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nde Paula Freitas B, Zin A, Ko A, et al.: Anterior-Segment Ocular Findings and Microphthalmia in Congenital Zika Syndrome. Ophthalmology. 2017; 124(12): 1876–8. PubMed Abstract \| Publisher Full Text\n\nLinden VV, Linden HV Junior, Leal MC, et al.: Discordant clinical outcomes of congenital Zika virus infection in twin pregnancies. Arq Neuropsiquiatr. 2017; 75(6): 381–6. PubMed Abstract \| Publisher Full Text\n\nLeal MC, van der Linden V, Bezerra TP, et al.: Characteristics of Dysphagia in Infants with Microcephaly Caused by Congenital Zika Virus Infection, Brazil, 2015. Emerg Infect Dis. 2017; 23(8): 1253–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nParra-Saavedra M, Reefhuis J, Piraquive JP, et al.: Serial Head and Brain Imaging of 17 Fetuses With Confirmed Zika Virus Infection in Colombia, South America. Obstet Gynecol. 2017; 130(1): 207–12. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAragao MFVV, Holanda AC, Brainer-Lima AM, et al.: Nonmicrocephalic Infants with Congenital Zika Syndrome Suspected Only after Neuroimaging Evaluation Compared with Those with Microcephaly at Birth and Postnatally: How Large Is the Zika Virus \"Iceberg\"? AJNR Am J Neuroradiol. 2017; 38(7): 1427–34. PubMed Abstract \| Publisher Full Text\n\nCabral CM, Nóbrega M, Leite PLE, et al.: Clinical-epidemiological description of live births with microcephaly in the state of Sergipe, Brazil, 2015. Epidemiol Serv Saude. 2017; 26(2): 245–54. PubMed Abstract \| Publisher Full Text\n\nSousa AQ, Cavalcante DIM, Franco LM, et al.: Postmortem Findings for 7 Neonates with Congenital Zika Virus Infection. Emerg Infect Dis. 2017; 23(7): 1164–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVentura LO, Ventura CV, Lawrence L, et al.: Visual impairment in children with congenital Zika syndrome. J AAPOS. 2017; 21(4): 295–299.e2. PubMed Abstract \| Publisher Full Text\n\nRamalho FS, Yamamoto AY, da Silva LL, et al.: Congenital Zika Virus Infection Induces Severe Spinal Cord Injury. Clin Infect Dis. 2017; 65(4): 687–90. PubMed Abstract \| Publisher Full Text\n\nCavalcanti DD, Alves LV, Furtado GJ, et al.: Echocardiographic findings in infants with presumed congenital Zika syndrome: Retrospective case series study. PLoS One. 2017; 12(4): e0175065. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYepez JB, Murati FA, Pettito M, et al.: Ophthalmic Manifestations of Congenital Zika Syndrome in Colombia and Venezuela. JAMA Ophthalmol. 2017; 135(5): 440–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAragao MFVV, Brainer-Lima AM, Holanda AC, et al.: Spectrum of Spinal Cord, Spinal Root, and Brain MRI Abnormalities in Congenital Zika Syndrome with and without Arthrogryposis. AJNR Am J Neuroradiol. 2017; 38(5): 1045–53. PubMed Abstract \| Publisher Full Text\n\nChimelli L, Melo ASO, Avvad-Portari E, et al.: The spectrum of neuropathological changes associated with congenital Zika virus infection. Acta Neuropathol. 2017; 133(6): 983–99. PubMed Abstract \| Publisher Full Text\n\nMeneses JDA, Ishigami AC, de Mello LM, et al.: Lessons Learned at the Epicenter of Brazil's Congenital Zika Epidemic: Evidence From 87 Confirmed Cases. Clin Infect Dis. 2017; 64(10): 1302–8. PubMed Abstract \| Publisher Full Text\n\nDel Campo M, Feitosa IM, Ribeiro EM, et al.: The phenotypic spectrum of congenital Zika syndrome. Am J Med Genet A. 2017; 173(4): 841–57. PubMed Abstract \| Publisher Full Text\n\nAcosta-Reyes J, Navarro E, Herrera MJ, et al.: Severe Neurologic Disorders in 2 Fetuses with Zika Virus Infection, Colombia. Emerg Infect Dis. 2017; 23(6): 982–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSanín-Blair JE, Gutiérrez-Márquez C, Herrera DA, et al.: Fetal Magnetic Resonance Imaging Findings in Prenatal Zika Virus Infection. Fetal Diagn Ther. 2017; 42(2): 153–7. PubMed Abstract \| Publisher Full Text\n\nSchaub B, Vouga M, Najioullah F, et al.: Analysis of blood from Zika virus-infected fetuses: a prospective case series. Lancet Infect Dis. 2017; 17(5): 520–7. PubMed Abstract \| Publisher Full Text\n\nTse C, Picon M, Rodriguez P, et al.: The Effects of Zika in Pregnancy: The Miami Experience [20M]. Obstet Gynecol. 2017; 129: 137s-8s. Publisher Full Text\n\nAleman TS, Ventura CV, Cavalcanti MM, et al.: Quantitative Assessment of Microstructural Changes of the Retina in Infants With Congenital Zika Syndrome. JAMA Ophthalmol. 2017; 135(10): 1069–76. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMejdoubi M, Monthieux A, Cassan T, et al.: Brain MRI in Infants after Maternal Zika Virus Infection during Pregnancy. N Engl J Med. 2017; 377(14): 1399–400. PubMed Abstract \| Publisher Full Text\n\nFernandez MP, Parra Saad E, Ospina Martinez M, et al.: Ocular Histopathologic Features of Congenital Zika Syndrome. JAMA Ophthalmol. 2017; 135(11): 1163–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPetribu NCL, Aragao MFV, van der Linden V, et al.: Follow-up brain imaging of 37 children with congenital Zika syndrome: case series study. BMJ. 2017; 359: j4188. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchaub B, Gueneret M, Jolivet E, et al.: Ultrasound imaging for identification of cerebral damage in congenital Zika virus syndrome: a case series. Lancet Child Adolesc Health. 2017; 1(1): 45–55. PubMed Abstract \| Publisher Full Text\n\nJames-Powell T, Brown Y, Christie CDC, et al.: Trends of Microcephaly and Severe Arthrogryposis in Three Urban Hospitals following the Zika, Chikungunya and Dengue Fever Epidemics of 2016 in Jamaica. West Indian Med J. 2017; 66(1): 10–9. Publisher Full Text\n\nContreras-Capetillo SN, Valadéz-González N, Manrique-Saide P, et al.: Birth Defects Associated With Congenital Zika Virus Infection in Mexico. Clin Pediatr (Phila). 2018; 57(8): 927–36. PubMed Abstract \| Publisher Full Text\n\nCastro JDV, Pereira LP, Dias DA, et al.: Presumed Zika virus-related congenital brain malformations: the spectrum of CT and MRI findings in fetuses and newborns. Arq Neuropsiquiatr. 2017; 75(10): 703–10. PubMed Abstract \| Publisher Full Text\n\nMulkey SB, Vezina G, Bulas DI, et al.: Neuroimaging Findings in Normocephalic Newborns With Intrauterine Zika Virus Exposure. Pediatr Neurol. 2018; 78: 75–8. PubMed Abstract \| Publisher Full Text\n\nPires P, Jungmann P, Galvão JM, et al.: Neuroimaging findings associated with congenital Zika virus syndrome: case series at the time of first epidemic outbreak in Pernambuco State, Brazil. Childs Nerv Syst. 2018; 34(5): 957–63. PubMed Abstract \| Publisher Full Text\n\nFelix A, Hallet E, Favre A, et al.: Cerebral injuries associated with Zika virus in utero exposure in children without birth defects in French Guiana: Case report. Medicine (Baltimore). 2017; 96(51): e9178. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRamos CL, Moreno-Carvalho OA, Nascimento-Carvalho CM: Cerebrospinal fluid aspects of neonates with or without microcephaly born to mothers with gestational Zika virus clinical symptoms. J Infect. 2018; 76(6): 563–9. PubMed Abstract \| Publisher Full Text\n\nSoares-Marangoni DA, Tedesco NM, Nascimento AL, et al.: General movements and motor outcomes in two infants exposed to Zika virus: brief report. Dev Neurorehabil. 2019; 22(1): 71–4. PubMed Abstract \| Publisher Full Text\n\nHoward A, Visintine J, Fergie J, et al.: Two Infants with Presumed Congenital Zika Syndrome, Brownsville, Texas, USA, 2016-2017. Emerg Infect Dis. 2018; 24(4): 625–30. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDe Fatima Viana Vasco Aragao M, Van Der Linden V, Petribu NC, et al.: Congenital Zika Syndrome: Comparison of brain CT scan with postmortem histological sections from the same subjects. Neuroradiology. 2018; 60(1 Supplement 1): 367–8.\n\nRosenstierne MW, Schaltz-Buchholzer F, Bruzadelli F, et al.: Zika Virus IgG in Infants with Microcephaly, Guinea-Bissau, 2016. Emerg Infect Dis. 2018; 24(5): 948–50. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOliveira-Filho J, Felzemburgh R, Costa F, et al.: Seizures as a Complication of Congenital Zika Syndrome in Early Infancy. Am J Trop Med Hyg. 2018; 98(6): 1860–2. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWalker CL, Merriam AA, Ohuma EO, et al.: Femur-sparing pattern of abnormal fetal growth in pregnant women from New York City after maternal Zika virus infection. Am J Obstet Gynecol. 2018; 219(2): 187.e1–187.e20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPeixoto Filho AAA, de Freitas SB, Ciosaki MM, et al.: Computed tomography and magnetic resonance imaging findings in infants with microcephaly potentially related to congenital Zika virus infection. Radiol Bras. 2018; 51(2): 119–22. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDe Alcantara T, La Beaud AD, Aronoff D, et al.: CT Scan Findings in Microcephaly Cases During 2015–2016 Zika Outbreak: A Cohort Study. Neurology. 2018; 90(15 Supplement 1). Reference Source\n\nCardoso TF Jr, Santos RSD, Corrêa RM, et al.: Congenital Zika infection: neurology can occur without microcephaly. Arch Dis Child. 2019; 104(2): 199–200. PubMed Abstract \| Publisher Full Text\n\nAlves LV, Mello MJG, Bezerra PG, et al.: Congenital Zika Syndrome and Infantile Spasms: Case Series Study. J Child Neurol. 2018; 33(10): 664–6. PubMed Abstract \| Publisher Full Text\n\nDe Moraes CG, Pettito M, Yepez JB, et al.: Optic neuropathy and congenital glaucoma associated with probable Zika virus infection in Venezuelan patients. JMM Case Rep. 2018; 5(5): e005145. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFigueredo LF, Franco-Zuluaga JA, Carrere-Rivera C, et al.: Anatomical Findings of fetuses Vertically-infected with Zika Virus. FASEB J. 2018; 32(1). Reference Source\n\nDe Alcantara T, Da Silva Maia JT, Aronoff D, et al.: Radiological findings and neurological disorders in microcephaly cases related to zika virus: A cohort study. Neurology. 2018; 90(15 Supplement 1). Reference Source\n\nAlmeida IMLM, Ramos CV, Rodrigues DC, et al.: Clinical and epidemiological aspects of microcephaly in the state of Piauí, northeastern Brazil, 2015-2016. J Pediatr (Rio J). 2019; 95(4): 466–474. PubMed Abstract \| Publisher Full Text\n\nWongsurawat T, Athipanyasilp N, Jenjaroenpun P, et al.: Case of Microcephaly after Congenital Infection with Asian Lineage Zika Virus, Thailand. Emerg Infect Dis. 2018; 24(9): 1758–1761. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRajapakse NS, Ellsworth K, Liesman RM, et al.: Unilateral Phrenic Nerve Palsy in Infants with Congenital Zika Syndrome. Emerg Infect Dis. 2018; 24(8): 1422–1427. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBeaufrere A, Bessieres B, Bonniere M, et al.: A clinical and histopathological study of malformations observed in fetuses infected by the Zika virus. Brain Pathol. 2019; 29(1): 114–125. PubMed Abstract \| Publisher Full Text\n\nChu V, Petersen LR, Moore CA, et al.: Possible Congenital Zika Syndrome in Older Children Due to Earlier Circulation of Zika Virus. Am J Med Genet A. 2018; 176(9): 1882–9. PubMed Abstract \| Publisher Full Text\n\nVouga M, Baud D, Jolivet E, et al.: Congenital Zika virus syndrome...what else? Two case reports of severe combined fetal pathologies. BMC Pregnancy Childbirth. 2018; 18(1): 356. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nde Noronha L, Zanluca C, Burger M, et al.: Zika Virus Infection at Different Pregnancy Stages: Anatomopathological Findings, Target Cells and Viral Persistence in Placental Tissues. Front Microbiol. 2018; 9: 2266. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMartins RS, Froes MH, Saad LDC, et al.: Descriptive report of cases of congenital syndrome associated with Zika virus infection in the state of São Paulo, Brazil, from 2015 to 2017. Epidemiol Serv Saude. 2018; 27(3): e2017382. PubMed Abstract \| Publisher Full Text\n\nRodriguez J, Diaz M, Montalvo A, et al.: Case Series: Congenital Zika Virus Infection Associated with Epileptic Spasms. Ann Neurol. 2018; 84(Supplement 22): S401–S. Publisher Full Text\n\nAzevedo RSS, Araujo MT, Oliveira CS, et al.: Zika Virus Epidemic in Brazil. II. Post-Mortem Analyses of Neonates with Microcephaly, Stillbirths, and Miscarriage. J Clin Med. 2018; 7(12): pii: E496. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPetribu NCL, Fernandes ACV, Abath MB, et al.: Common findings on head computed tomography in neonates with confirmed congenital Zika syndrome. Radiol Bras. 2018; 51(6): 366–71. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSeferovic MD, Turley M, Valentine GC, et al.: Clinical Importance of Placental Testing among Suspected Cases of Congenital Zika Syndrome. Int J Mol Sci. 2019; 20(3): pii: E712. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCano M, Esquivel R: Zika virus infection in Hospital del Niño 'Dr José Renán Esquivel' (Panamá): Case Review since the introduction in Latin America. Pediátr Panamá. 2018; 47(3): 15–9. Reference Source\n\nde Fatima Viana Vasco Aragao M, van der Linden V, Petribu NC, et al.: Congenital Zika Syndrome: The Main Cause of Death and Correspondence Between Brain CT and Postmortem Histological Section Findings From the Same Individuals. Top Magn Reson Imaging. 2019; 28(1): 29–33. PubMed Abstract \| Publisher Full Text\n\nSantana MB, Lamas CC, Athayde JG, et al.: Congenital Zika syndrome: is the heart part of its spectrum? Clin Microbiol Infect. 2019; 25(8): 1043–4. PubMed Abstract \| Publisher Full Text\n\nVenturi G, Fortuna C, Alves RM, et al.: Epidemiological and clinical suspicion of congenital Zika virus infection: Serological findings in mothers and children from Brazil. J Med Virol. 2019; 91(9): 1577–1583. PubMed Abstract \| Publisher Full Text\n\nLee EH, Cooper H, Iwamoto M, et al.: First 12 Months of Life for Infants in New York City, New York, With Possible Congenital Zika Virus Exposure. J Pediatric Infect Dis Soc. 2019; pii: piz027. PubMed Abstract \| Publisher Full Text\n\nMerriam AA, Nhan-Chang CL, Huerta-Bogdan BI, et al.: A Single-Center Experience with a Pregnant Immigrant Population and Zika Virus Serologic Screening in New York City. Am J Perinatol. 2019. PubMed Abstract \| Publisher Full Text\n\nZambrano H, Waggoner J, Leon K, et al.: High incidence of Zika virus infection detected in plasma and cervical cytology specimens from pregnant women in Guayaquil, Ecuador. Am J Reprod Immunol. 2017; 77(2): e12630. PubMed Abstract \| Publisher Full Text\n\nSanta Rita TH, Barra RB, Peixoto GP, et al.: Association between suspected Zika virus disease during pregnancy and giving birth to a newborn with congenital microcephaly: a matched case-control study. BMC Res Notes. 2017; 10(1): 457. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nde Araujo TVB, Ximenes RAA, Miranda-Filho DB, et al.: Association between microcephaly, Zika virus infection, and other risk factors in Brazil: final report of a case-control study. Lancet Infect Dis. 2018; 18(3): 328–36. PubMed Abstract \| Publisher Full Text\n\nMoreira-Soto A, Sarno M, Pedroso C, et al.: Evidence for Congenital Zika Virus Infection From Neutralizing Antibody Titers in Maternal Sera, Northeastern Brazil. J Infect Dis. 2017; 216(12): 1501–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKrow-Lucal ER, de Andrade MR, Cananéa JNA, Moore CA, et al.: Association and birth prevalence of microcephaly attributable to Zika virus infection among infants in Paraíba Brazil, in 201 6: a case-control study. The Lancet Child & Adolescent Health. 2018; 2(3): 205–13. Publisher Full Text\n\nVentura LO, Ventura CV, Dias NC, et al.: Visual impairment evaluation in 119 children with congenital Zika syndrome. J AAPOS. 2018; 22(3): 218–22 e1. PubMed Abstract \| Publisher Full Text\n\nMoreira-Soto A, Cabral R, Pedroso C, et al.: Exhaustive TORCH Pathogen Diagnostics Corroborate Zika Virus Etiology of Congenital Malformations in Northeastern Brazil. mSphere. 2018; 3(4): pii: e00278-18. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSubissi L, Dub T, Besnard M, et al.: Zika Virus Infection during Pregnancy and Effects on Early Childhood Development, French Polynesia, 2013-2016. Emerg Infect Dis. 2018; 24(10): 1850–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLima GP, Rozenbaum D, Pimentel C, et al.: Factors associated with the development of Congenital Zika Syndrome: a case-control study. BMC Infect Dis. 2019; 19(1): 277. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPedroso C, Fischer C, Feldmann M, et al.: Cross-Protection of Dengue Virus Infection against Congenital Zika Syndrome, Northeastern Brazil. Emerg Infect Dis. 2019; 25(8): 1485–1493. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZin AA, Tsui I, Rossetto J, et al.: Screening Criteria for Ophthalmic Manifestations of Congenital Zika Virus Infection. JAMA Pediatr. 2017; 171(9): 847–54. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVercosa I, Carneiro P, Vercosa R, et al.: The visual system in infants with microcephaly related to presumed congenital Zika syndrome. J AAPOS. 2017; 21(4): 300–4 e1. PubMed Abstract \| Publisher Full Text\n\nHalai UA, Nielsen-Saines K, Moreira ML, et al.: Maternal Zika Virus Disease Severity, Virus Load, Prior Dengue Antibodies, and Their Relationship to Birth Outcomes. Clin Infect Dis. 2017; 65(6): 877–83. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nReynolds MR, Jones AM, Petersen EE, et al.: Vital Signs: Update on Zika Virus-Associated Birth Defects and Evaluation of All U.S. Infants with Congenital Zika Virus Exposure - U.S. Zika Pregnancy Registry, 2016. MMWR Morb Mortal Wkly Rep. 2017; 66(13): 366–73. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAdhikari EH, Nelson DB, Johnson KA, et al.: Infant outcomes among women with Zika virus infection during pregnancy: results of a large prenatal Zika screening program. Am J Obstet Gynecol. 2017; 216(3): 292 e1–e8. PubMed Abstract \| Publisher Full Text\n\nSpiliopoulos D, Wooton G, Economides DL, et al.: Surveillance of pregnant women exposed to Zika virus areas. Bjog-Int J Obstet Gy. 2017; 124: 17–49. Publisher Full Text\n\nEppes C, Rac M, Dempster C, et al.: Zika Virus in a Non-Endemic Urban Population: Patient Characteristics and Ultrasound Findings. Obstet Gynecol. 2017; 129: 135s-s. Publisher Full Text\n\nSohan K, Cyrus CA: Ultrasonographic observations of the fetal brain in the first 100 pregnant women with Zika virus infection in Trinidad and Tobago. Int J Gynaecol Obstet. 2017; 139(3): 278–83. PubMed Abstract \| Publisher Full Text\n\nKam YW, Leite JA, Lum FM, et al.: Specific Biomarkers Associated With Neurological Complications and Congenital Central Nervous System Abnormalities From Zika Virus-Infected Patients in Brazil. J Infect Dis. 2017; 216(2): 172–81. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nXimenes ASFC, Pires P, Werner H, et al.: Neuroimaging findings using transfontanellar ultrasound in newborns with microcephaly: a possible association with congenital Zika virus infection. J Matern Fetal Neonatal Med. 2019; 32(3): 493–501. PubMed Abstract \| Publisher Full Text\n\nTerzian ACB, Estofolete CF, Alves da Silva R, et al.: Long-Term Viruria in Zika Virus-Infected Pregnant Women, Brazil, 2016. Emerg Infect Dis. 2017; 23(11): 1891–3. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNogueira ML, Nery Junior NRR, Estofolete CF, et al.: Adverse birth outcomes associated with Zika virus exposure during pregnancy in São José do Rio Preto, Brazil. Clin Microbiol Infect. 2018; 24(6): 646–52. PubMed Abstract \| Publisher Full Text\n\nSanz Cortes M, Rivera AM, Yepez M, et al.: Clinical assessment and brain findings in a cohort of mothers, fetuses and infants infected with ZIKA virus. Am J Obstet Gynecol. 2018; 218(4): 440 e1–e36. PubMed Abstract \| Publisher Full Text\n\nMaykin M, Avaad-Portari E, Esquivel M, et al.: Placental Histopathologic Findings in Zika-infected Pregnancies. Am J Obstet Gynecol. 2018; 218(1): S520–S1. Publisher Full Text\n\nAdhikari EH, McKiever M, Nelson DB, et al.: Neonatal surveillance among asymptomatic Zika-exposed infants through 6 months of life. Am J Obstet Gynecol. 2018; 218(1): S512–S3. Publisher Full Text\n\nHoen B, Schaub B, Funk AL, et al.: Pregnancy Outcomes after ZIKV Infection in French Territories in the Americas. N Engl J Med. 2018; 378(11): 985–94. PubMed Abstract \| Publisher Full Text\n\nCollier ARY, Barouch DH: Maternal Immune Response to Zika Virus. Reproductive Sciences. 2018; 25(1): 163a-4a.\n\nBulas D, Mulkey S, Vezina G, et al.: Fetal and postnatal brain imaging for the detection of ZIKV encephalopathy in the fetus/newborn. Pediatric Radiology. 2018; 48(1 Supplement 1): S141.\n\nRodriguez-Morales AJ, Cardona-Ospina JA, Ramirez-Jaramillo V, et al.: Diagnosis and outcomes of pregnant women with Zika virus infection in two municipalities of Risaralda, Colombia: Second report of the ZIKERNCOL study. Travel Med Infect Dis. 2018; 25: 20–5. PubMed Abstract \| Publisher Full Text\n\nRice ME, Galang RR, Roth NM, et al.: Vital Signs: Zika-Associated Birth Defects and Neurodevelopmental Abnormalities Possibly Associated with Congenital Zika Virus Infection - U.S. Territories and Freely Associated States, 2018. MMWR Morb Mortal Wkly Rep. 2018; 67(31): 858–67. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTsui I, Moreira MEL, Rossetto JD, et al.: Eye Findings in Infants With Suspected or Confirmed Antenatal Zika Virus Exposure. Pediatrics. 2018; 142(4): pii: e20181104. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDávila-Camargo A, Durán-Nah JJ, Dupinet-Sánchez A: Ophthalmologic findings in newborns of women with Zika virus infection during pregnancy: A case series. Revista Mexicana de Oftalmología. (English Edition) 2018; 92(4): 159–62. Publisher Full Text\n\nPomar L, Vouga M, Lambert V, et al.: Maternal-fetal transmission and adverse perinatal outcomes in pregnant women infected with Zika virus: prospective cohort study in French Guiana. BMJ. 2018; 363: k4431. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrock M, Magalhaes AQ, Di Tommaso Leao J, et al.: Zika virus infection in pregnant women in Manaus. Int J Gynaecol Obstet. 2018; 143(Supplement 3): 725–6. Publisher Full Text\n\nCaixeta R, Magalhaes AC, De Brito WI, et al.: Vertical transmission of zika virus and perinatal microcephaly. Int J Gynaecol Obstet. 2018; 143(Supplement 3): 725. Publisher Full Text\n\nRosado L, Gomes MBF, Lacerda FA, et al.: Cohort of infected Zika Virus pregnant women from central Brazil: Preliminary report. Int J Gynaecol Obstet. 2018; 143(Supplement 3): 228. Publisher Full Text\n\nMulkey SB, Bulas DI, Vezina G, et al.: Sequential Neuroimaging of the Fetus and Newborn With In Utero Zika Virus Exposure. JAMA Pediatr. 2019; 173(1): 52–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGely-Rojas L, Pérez R, García-Fragoso L, et al.: Association of Zika Virus Exposure in Utero with Ocular Phenotypes in a Group of Newborns in Puerto Rico Exposure with ocular phenotypes in newborns in Puerto Rico. P R Health Sci J. 2018; 37(Special Issue): S77–S80. PubMed Abstract\n\nGely-Rojas L, García-Fragoso L, Negrón J, et al.: Congenital Zika Syndrome in Puerto Rico, Beyond Microcephaly, A Multiorgan Approach. P R Health Sci J. 2018; 37(Special Issue): S73–S6. PubMed Abstract\n\nPereira JP Jr, Maykin MM, Vasconcelos Z, et al.: The Role of Amniocentesis in the Diagnosis of Congenital Zika Syndrome. Clin Infect Dis. 2019; 69(4): 713–716. PubMed Abstract \| Publisher Full Text\n\nPereira JP Jr, Nielsen-Saines K, Sperling J, et al.: Association of Prenatal Ultrasonographic Findings With Adverse Neonatal Outcomes Among Pregnant Women With Zika Virus Infection in Brazil. JAMA Netw Open. 2018; 1(8): e186529. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDel Carpio-Orantes L, Rosas-Lozano AL, García-Méndez S: Zika virus infection in pregnant women in a General Hospital of Veracruz, Mexico. J Matern Fetal Neonatal Med. 2019; 1–5. PubMed Abstract \| Publisher Full Text\n\nRodó C, Suy A, Sulleiro E, et al.: Pregnancy outcomes after maternal Zika virus infection in a non-endemic region: prospective cohort study. Clin Microbiol Infect. 2019; 25(5): 633.e5–633.e9. PubMed Abstract \| Publisher Full Text\n\nCalle-Giraldo JP, Rojas CA, Hurtado IC, et al.: Outcomes of Congenital Zika Virus Infection During an Outbreak in Valle del Cauca, Colombia. Pediatr Infect Dis J. 2019; 38(7): 735–40. PubMed Abstract \| Publisher Full Text\n\nBaran LCP, da Costa MF, Vidal KS, et al.: Alterations in visual acuity and visual development in infants 1-24 months old either exposed to or infected by Zika virus during gestation, with and without microcephaly. J AAPOS. 2019; pii: S1091-8531(19)30137-5. PubMed Abstract \| Publisher Full Text\n\nde Magalhães-Barbosa MC, Prata-Barbosa A, Robaina JR, et al.: Prevalence of microcephaly in eight south-eastern and midwestern Brazilian neonatal intensive care units: 2011-2015. Arch Dis Child. 2017; 102(8): 728–34. PubMed Abstract \| Publisher Full Text\n\nMerriam AA, Nhan-Chang CL, Huerta-Bogdan BI, et al.: 393: A single-center experience with a pregnant immigrant population and zika virus in new york city. Am J Perinatol. 2017; 216(1): S236-S. Publisher Full Text\n\nGregianini TS, Ranieri T, Favreto C, et al.: Emerging arboviruses in Rio Grande do Sul, Brazil: Chikungunya and Zika outbreaks, 2014-2016. Rev Med Virol. 2017; 27(6). PubMed Abstract \| Publisher Full Text\n\nNetto EM, Moreira-Soto A, Pedroso C, et al.: High Zika Virus Seroprevalence in Salvador, Northeastern Brazil Limits the Potential for Further Outbreaks. MBio. 2017; 8(6): pii: e01390-17. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShiu C, Starker R, Kwal J, et al.: Zika Virus Testing and Outcomes during Pregnancy, Florida, USA, 2016. Emerg Infect Dis. 2018; 24(1): 1–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMletzko J, Schildgen O: Absence of Zika virus in abortion and placental tissue in a German cohort. Rev Med Microbiol. 2018; 29(1): 17–9. Reference Source\n\nMarques Abramov D, Saad T, Gomes-Junior SC, et al.: Auditory brainstem function in microcephaly related to Zika virus infection. Neurology. 2018; 90(7): e606–e14. PubMed Abstract \| Publisher Full Text\n\nOrofino DHG, Passos SRL, de Oliveira RVC, et al.: Cardiac findings in infants with in utero exposure to Zika virus- a cross sectional study. PLoS Negl Trop Dis. 2018; 12(3): e0006362. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFonteles CSR, Marques Ribeiro E, Sales Aragão Santos M, et al.: Lingual Frenulum Phenotypes in Brazilian Infants With Congenital Zika Syndrome. Cleft Palate Craniofac J. 2018; 55(10): 1391–8. PubMed Abstract \| Publisher Full Text\n\nFerreira HNC, Schiariti V, Regalado ICR, et al.: Functioning and Disability Profile of Children with Microcephaly Associated with Congenital Zika Virus Infection. Int J Environ Res Public Health. 2018; 15(6): pii: E1107. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDa Maia JTS, Vidal SJA, Mendes TV, et al.: Perinatal Case Fatality Rate in Cases of Congenital Zika Syndrome: a Cross-Sectional Study. Neurology. 2018; 90(15 Supplement 1). Reference Source\n\nJoão EC, Ferreira ODC Jr, Gouvêa MI, et al.: Pregnant women co-infected with HIV and Zika: Outcomes and birth defects in infants according to maternal symptomatology. PLoS One. 2018; 13(7): e0200168. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHerber S, Silva AA, Sanseverino MTV, et al.: Prevalence and causes of congenital microcephaly in the absence of a Zika virus outbreak in southern Brazil. J Pediatr (Rio J). 2018; pii: S0021-7557(18)30119-0. PubMed Abstract \| Publisher Full Text\n\nZin AA, Tsui I, Rossetto JD, et al.: Visual function in infants with antenatal Zika virus exposure. J AAPOS. 2018; 22(6): 452–456.e1. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKikuti M, Cardoso CW, Prates APB, et al.: Congenital brain abnormalities during a Zika virus epidemic in Salvador, Brazil, April 2015 to July 2016. Euro Surveill. 2018; 23(45): 1700757. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAlayed MS, Qureshi MA, Ahmed S, et al.: Seroprevalence of Zika virus among asymptomatic pregnant mothers and their newborns in the Najran region of southwest Saudi Arabia. Ann Saudi Med. 2018; 38(6): 408–12. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPeña F, Pimentel R, Khosla S, et al.: Zika Virus Epidemic in Pregnant Women, Dominican Republic, 2016-2017. Emerg Infect Dis. 2019; 25(2): 247–55. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nC Lage ML, Carvalho AL, Ventura PA, et al.: Clinical, Neuroimaging, and Neurophysiological Findings in Children with Microcephaly Related to Congenital Zika Virus Infection. Int J Environ Res Public Health. 2019; 16(3): pii: E309. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCarvalho FR, Medeiros T, Vianna RAO, et al.: Simultaneous circulation of arboviruses and other congenital infections in pregnant women in Rio de Janeiro, Brazil. Acta Trop. 2019; 192(NA): 49–54. PubMed Abstract \| Publisher Full Text\n\nJaenisch T, Rosenberger KD, Brito C, et al.: Risk of microcephaly after Zika virus infection in Brazil, 2015 to 2016. Bull World Health Organ. 2017; 95(3): 191–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSouza WV, Albuquerque MFPM, Vazquez E, et al.: Microcephaly epidemic related to the Zika virus and living conditions in Recife, Northeast Brazil. BMC Public Health. 2018; 18(1): 130. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVissoci JRN, Rocha TAH, Silva NCD, et al.: Zika virus infection and microcephaly: Evidence regarding geospatial associations. PLoS Negl Trop Dis. 2018; 12(4): e0006392. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSouza AI, de Siqueira MT, Ferreira ALCG, et al.: Geography of Microcephaly in the Zika Era: A Study of Newborn Distribution and Socio-environmental Indicators in Recife, Brazil, 2015-2016. Public Health Rep. 2018; 133(4): 461–71. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMajumder MS, Hess R, Ross R, et al.: Seasonality of birth defects in West Africa: could congenital Zika syndrome be to blame? [version 2; peer review: 2 approved, 1 approved with reservations] F1000Res. 2018; 7: 159. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCampos MC, Dombrowski JG, Phelan J, et al.: Zika might not be acting alone: Using an ecological study approach to investigate potential co-acting risk factors for an unusual pattern of microcephaly in Brazil. PLoS One. 2018; 13(8): e0201452. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrady OJ, Osgood-Zimmerman A, Kassebaum NJ, et al.: The association between Zika virus infection and microcephaly in Brazil 2015-2017: An observational analysis of over 4 million births. PLoS Med. 2019; 16(3): e1002755. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBautista LE: Maternal Zika virus infection and newborn microcephaly-an analysis of the epidemiological evidence. Ann Epidemiol. 2018; 28(2): 111–8. PubMed Abstract \| Publisher Full Text\n\nHernández-Ávila JE, Palacio-Mejía LS, López-Gatell H, et al.: Zika virus infection estimates, Mexico. Bull World Health Organ. 2018; 96(5): 306–13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHay JA, Nouvellet P, Donnelly CA, et al.: Potential inconsistencies in Zika surveillance data and our understanding of risk during pregnancy. PLoS Negl Trop Dis. 2018; 12(12): e0006991. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHurtado-Villa P, Puerto AK, Victoria S, et al.: Raised Frequency of Microcephaly Related to Zika Virus Infection in Two Birth Defects Surveillance Systems in Bogotá and Cali, Colombia. Pediatr Infect Dis J. 2017; 36(10): 1017–9. PubMed Abstract \| Publisher Full Text\n\nde Oliveira WK, de França GVA, Carmo EH, et al.: Infection-related microcephaly after the 2015 and 2016 Zika virus outbreaks in Brazil: a surveillance-based analysis. Lancet. 2017; 390(10097): 861–70. PubMed Abstract \| Publisher Full Text\n\nShapiro-Mendoza CK, Rice ME, Galang RR, et al.: Pregnancy Outcomes After Maternal Zika Virus Infection During Pregnancy - U.S. Territories, January 1, 2016-April 25, 2017. MMWR Morb Mortal Wkly Rep. 2017; 66(23): 615–21. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nde Oliveira WK, Carmo EH, Henriques CM, et al.: Zika Virus Infection and Associated Neurologic Disorders in Brazil. N Engl J Med. 2017; 376(16): 1591–3. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMagalhães-Barbosa MC, Prata-Barbosa A, Robaina JR, et al.: New trends of the microcephaly and Zika virus outbreak in Brazil, July 2016-December 2016. Travel Med Infect Dis. 2017; 16: 52–7. PubMed Abstract \| Publisher Full Text\n\nJournel I, Andrécy LL, Metellus D, et al.: Transmission of Zika Virus - Haiti, October 12, 2015-September 10, 2016. MMWR Morb Mortal Wkly Rep. 2017; 66(6): 172–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNeto NNM, Da Silva Maia JT, Zacarkim MR, et al.: Mortality in newborns with microcephaly due to maternal zika virus infection in Rio Grande Do Norte state-northeast Brazil: A crosssectional study. Neurology. 2017; 88(16 Supplement 1). Reference Source\n\nHall NB, Broussard K, Evert N, et al.: Notes from the Field: Zika Virus-Associated Neonatal Birth Defects Surveillance - Texas, January 2016-July 2017. MMWR Morb Mortal Wkly Rep. 2017; 66(31): 835–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMéndez N, Oviedo-Pastrana M, Mattar S, et al.: Zika virus disease, microcephaly and Guillain-Barré syndrome in Colombia: epidemiological situation during 21 months of the Zika virus outbreak, 2015-2017. Arch Public Health. 2017; 75: 65. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDelaney A, Mai C, Smoots A, et al.: Population-Based Surveillance of Birth Defects Potentially Related to Zika Virus Infection - 15 States and U.S. Territories, 2016. MMWR Morb Mortal Wkly Rep. 2018; 67(3): 91–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRibeiro IG, Andrade MR, Silva JM, et al.: Microcephaly in Piauí, Brazil: descriptive study during the Zika virus epidemic, 2015-2016. Epidemiol Serv Saude. 2018; 27(1): e20163692. PubMed Abstract \| Publisher Full Text\n\nMendes Neto NN, da Silva Maia JT, Zacarkim MR, et al.: Perinatal Case Fatality Rate Related to Congenital Zika Syndrome in Brazil: A Cross-Sectional Study. Pediatr Neurol. 2018; 81: 47–8. PubMed Abstract \| Publisher Full Text\n\nSowunmi S, Eckert V, Griffin A: 58th Annual Teratology Society Meeting. Birth Defects Res. 2018; 110(9): 667–822. PubMed Abstract \| Publisher Full Text\n\nTellechea AL, Luppo V, Morales MA, et al.: Surveillance of microcephaly and selected brain anomalies in Argentina: Relationship with Zika virus and other congenital infections. Birth Defects Res. 2018; 110(12): 1016–26. PubMed Abstract \| Publisher Full Text\n\nFrança GVA, Pedi VD, Garcia MHO, et al.: Congenital syndrome associated with Zika virus infection among live births in Brazil: a description of the distribution of reported and confirmed cases in 2015-2016. Epidemiol Serv Saude. 2018; 27(2): e2017473. PubMed Abstract \| Publisher Full Text\n\nPorse CC, Messenger S, Vugia DJ, et al.: Travel-Associated Zika Cases and Threat of Local Transmission during Global Outbreak, California, USA. Emerg Infect Dis. 2018; 24(9): 1626–32. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSilva JHD, Terças ACP, Pinheiro LCB, et al.: Profile of congenital anomalies among live births in the municipality of Tangará da Serra, Mato Grosso, Brazil, 2006-2016. Epidemiol Serv Saude. 2018; 27(3): e2018008. PubMed Abstract \| Publisher Full Text\n\nSpiliopoulos D, Panchal M, Economides DL: Surveillance of pregnant women with potential exposure to Zika virus following travel. G Chir. 2019; 40(1): 58–65. PubMed Abstract\n\nFernández Martínez B, Martínez Sánchez EV, Diaz Garcia O, et al.: Zika virus disease in Spain. Surveillance results and epidemiology on reported cases, 2015-2017. Med Clin (Barc). 2019; 153(1): 6–12. PubMed Abstract \| Publisher Full Text\n\nSulleiro E, Rando A, Alejo I, et al.: Screening for Zika virus infection in 1057 potentially exposed pregnant women, Catalonia (northeastern Spain). Travel Med Infect Dis. 2019; 29: 69–71. PubMed Abstract \| Publisher Full Text\n\nMiller E, Becker Z, Shalev D, et al.: Probable Zika virus-associated Guillain-Barré syndrome: Challenges with clinico-laboratory diagnosis. J Neurol Sci. 2017; 375: 367–70. PubMed Abstract \| Publisher Full Text\n\nRaboni SM, Bonfim C, Almeida BM, et al.: Flavivirus cross-reactivity in serological tests and Guillain-Barré syndrome in a hematopoietic stem cell transplant patient: A case report. Transpl Infect Dis. 2017; 19(4). PubMed Abstract \| Publisher Full Text\n\nTantillo G, Sclar G, Vasa C, et al.: Zika associated Guillain-Barre syndrome in the United States. Neurology. 2017; 88(16 Supplement 1). Reference Source\n\nSantos P, Nogueira F, Modenezi L, et al.: Movement disorder development later after acute disseminated encephalomyelitis associated with Zika virus: Case report. Neurology. 2017; 88(16 Supplement 1). Reference Source\n\nBeattie J, Parajuli S, Sanger M, et al.: Zika Virus-Associated Guillain-Barre Syndrome In A Returning United States Traveler. Am J Respir Crit Care Med. 2017; 195: A2008. Reference Source\n\nWu Y, Cui X, Wu N, et al.: A unique case of human Zika virus infection in association with severe liver injury and coagulation disorders. Sci Rep. 2017; 7(1): 11393. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeonhard SE, Munts AG, van der Eijk AA, et al.: Acute-onset chronic inflammatory demyelinating polyneuropathy after Zika virus infection. J Neurol Neurosurg Psychiatry. 2017; 89(10): 1118–1119. PubMed Abstract \| Publisher Full Text\n\nDirlikov E, Torres JV, Martines RB, et al.: Postmortem Findings in Patient with Guillain-Barré Syndrome and Zika Virus Infection. Emerg Infect Dis. 2018; 24(1): 114–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSiwarit R: Guillain-Barre syndrome in the Context of Zika Virus Infection: the First Report Case in Thailand. 2017; 34(2). Reference Source\n\nGibelin N, Romero J, Gregoire V, et al.: Miller Fisher syndrome associated with a Zika virus infection. Eur J Neurol. 2018; 25(2): e20–e1. PubMed Abstract \| Publisher Full Text\n\nGonzalez-Escobar G, Valadere AM, Adams R, et al.: Prolonged Zika virus viremia in a patient with Guillain-Barré syndrome in Trinidad and Tobago. Rev Panam Salud Publica. 2018; 41: e136. PubMed Abstract \| Publisher Full Text\n\nJones M: Multidisciplinary approach to Guillain-Barre Syndrome and neurologic injuries after zika virus. J Spinal Cord Med. 2018; 41(5): 588.\n\nMancera-Paez O, Román GC, Pardo-Turriago R, et al.: Concurrent Guillain-Barré syndrome, transverse myelitis and encephalitis post-Zika: A case report and review of the pathogenic role of multiple arboviral immunity. J Neurol Sci. 2018; 395: 47–53. PubMed Abstract \| Publisher Full Text\n\nCordeiro de Souza L, de Souza AA, de Almeida EEP, et al.: Inspiratory Muscle Training with Isokinetic Device to Help Ventilatory Weaning in a Patient with Guillain-Barré Syndrome by Zika Virus. Case Rep Crit Care. 2018; 2018: 9708451. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRivera-Concepción JR, Betancourt JP, Cerra J, et al.: The Zika Virus: An Association to Guillain-Barré Syndrome in the United States - A Case Report. P R Health Sci J. 2018; 37(Spec Issue): S93–S5. PubMed Abstract\n\nWright JK, Castellani L, Lecce C, et al.: Zika Virus-Associated Aseptic Meningitis and Guillain-Barre Syndrome in a Traveler Returning from Latin America: a Case Report and Mini-Review. Curr Infect Dis Rep. 2019; 21(1): 3. PubMed Abstract \| Publisher Full Text\n\nBoggild AK, Geduld J, Libman M, et al.: Surveillance report of Zika virus among Canadian travellers returning from the Americas. CMAJ. 2017; 189(9): E334–E40. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVillamil-Gomez WE, Sánchez-Herrera ÁR, Hernandez H, et al.: Guillain-Barré syndrome during the Zika virus outbreak in Sucre, Colombia, 2016. Travel Med Infect Dis. 2017; 16: 62–3. PubMed Abstract \| Publisher Full Text\n\nAli A, Williams M: Zik-V outbreak and Guillain-Barre syndrome in Jamaica. Neurology. 2017; 88(16 Supplement 1). Reference Source\n\nFrancois R, Berkowitz A: Zika-associated atypical guillain-barre variants in rural haiti. Neurology. 2017; 88(16 Supplement 1). Reference Source\n\nBarreira AA, Marques W, Marreco A, et al.: Guilllain-barre syndrome related to Zika virus infection in Brazil. Neurology. 2017; 88(16 Supplement 1). Reference Source\n\nBetances N, Luciano C, Carlo J, et al.: Prominent distal demyelination is a feature of zika-associated guillain barre syndrome. Neurology. 2017; 88(16 Supplement 1). Reference Source\n\nSebastián UU, Ricardo AVA, Alvarez BC, et al.: Zika virus-induced neurological critical illness in Latin America: Severe Guillain-Barre Syndrome and encephalitis. J Crit Care. 2017; 42: 275–81. PubMed Abstract \| Publisher Full Text Reference Source\n\nda Silva IRF, Frontera JA, Bispo de Filippis AM, et al.: Neurologic Complications Associated With the Zika Virus in Brazilian Adults. JAMA Neurol. 2017; 74(10): 1190–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDel Carpio Orantes L, Juárez Rangel FJ, García-Méndez S: Incidence of Guillain-Barré syndrome at a secondary centre during the 2016 zika outbreak. Neurologia. 2017; pii: S0213-4853(17)30279-7. PubMed Abstract \| Publisher Full Text\n\nDourado ME, Fernandes U, Vital AL, et al.: High Incidence of Guillain-Barre Syndrome after Zika Virus Infection in the State Rio Grande Do Norte, in Northeast Brazil. J Peripher Nerv Syst. 2017; 22(3): 275–6.\n\nRoze B, Najioullah F, Ferge JL, et al.: Guillain-Barré Syndrome Associated With Zika Virus Infection in Martinique in 2016: A Prospective Study. Clin Infect Dis. 2017; 65(9): 1462–8. PubMed Abstract \| Publisher Full Text\n\nBrito Ferreira ML, Antunes de Brito CA, Moreira ÁJP, et al.: Guillain-Barré Syndrome, Acute Disseminated Encephalomyelitis and Encephalitis Associated with Zika Virus Infection in Brazil: Detection of Viral RNA and Isolation of Virus during Late Infection. Am J Trop Med Hyg. 2017; 97(5): 1405–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nResiere D, Ferge JL, Fergé J, et al.: Cardiovascular complications in patients with Zika virus-induced Guillain-Barré syndrome. J Clin Virol. 2018; 98: 8–9. PubMed Abstract \| Publisher Full Text\n\nUncini A, Gonzalez-Bravo DC, Acosta-Ampudia YY, et al.: Clinical and nerve conduction features in Guillain-Barré syndrome associated with Zika virus infection in Cúcuta, Colombia. Eur J Neurol. 2018; 25(4): 644–50. PubMed Abstract \| Publisher Full Text\n\nMehta R, Soares CN, Medialdea-Carrera R, et al.: The spectrum of neurological disease associated with Zika and chikungunya viruses in adults in Rio de Janeiro, Brazil: A case series. PLoS Negl Trop Dis. 2018; 12(2): e0006212. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDirlikov E, Major CG, Medina NA, et al.: Clinical Features of Guillain-Barré Syndrome With vs Without Zika Virus Infection, Puerto Rico, 2016. JAMA Neurol. 2018; 75(9): 1089–97. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAzevedo MB, Coutinho MSC, Silva MAD, et al.: Neurologic manifestations in emerging arboviral diseases in Rio de Janeiro City, Brazil, 2015-2016. Rev Soc Bras Med Trop. 2018; 51(3): 347–51. PubMed Abstract \| Publisher Full Text\n\nNóbrega MEBD, Araujo ELL, Wada MY, et al.: Outbreak of Guillain-Barré syndrome possibly related to prior Zika virus infection, Metropolitan Region of Recife, Pernambuco, Brazil, 2015. Epidemiol Serv Saude. 2018; 27(2): e2017039. PubMed Abstract \| Publisher Full Text\n\nBaskar D, Amalnath D, Mandal J, et al.: Antibodies to Zika virus, Campylobacter jejuni and gangliosides in Guillain-Barre syndrome: A prospective single-center study from southern India. Neurol India. 2018; 66(5): 1324–31. PubMed Abstract \| Publisher Full Text\n\nZambrano LI, Fuentes-Barahona IC, Soto-Fernandez RJ, et al.: Guillain-Barré syndrome associated with Zika virus infection in Honduras, 2016-2017. Int J Infect Dis. 2019; 84: 136–7. PubMed Abstract \| Publisher Full Text\n\nSoto-Hernandez JL, Ponce de Leon Rosales S, Vargas Canas ES, et al.: Guillain-Barré Syndrome Associated With Zika Virus Infection: A Prospective Case Series From Mexico. Front Neurol. 2019; 10: 435. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStyczynski AR, Malta JMAS, Krow-Lucal ER, et al.: Increased rates of Guillain-Barré syndrome associated with Zika virus outbreak in the Salvador metropolitan area, Brazil. PLoS Negl Trop Dis. 2017; 11(8): e0005869. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSalinas JL, Walteros DM, Styczynski A, et al.: Zika virus disease-associated Guillain-Barré syndrome-Barranquilla, Colombia 2015-2016. J Neurol Sci. 2017; 381: 272–7. PubMed Abstract \| Publisher Full Text\n\nDirlikov E, Medina NA, Major CG, et al.: Acute Zika Virus Infection as a Risk Factor for Guillain-Barré Syndrome in Puerto Rico. JAMA. 2017; 318(15): 1498–500. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSimon O, Acket B, Forfait C, et al.: Zika virus outbreak in New Caledonia and Guillain-Barré syndrome: a case-control study. J Neurovirol. 2018; 24(3): 362–8. PubMed Abstract \| Publisher Full Text\n\nGeurtsvanKessel CH, Islam Z, Islam MB, et al.: Zika virus and Guillain-Barré syndrome in Bangladesh. Ann Clin Transl Neurol. 2018; 5(5): 606–15. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLynch RM, Mantus G, Encinales L, et al.: Augmented Zika and Dengue Neutralizing Antibodies Are Associated With Guillain-Barré Syndrome. J Infect Dis. 2019; 219(1): 26–30. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMunoz LS, Barréras P, Lizarazo J, et al.: Neuroviruses Emerging in the Americas Study(NEAS): The Colombian experience during the 2016 outbreak of Zika virus infection. Neurology. 2017; 88(16 Supplement 1). Reference Source\n\nBrito KGDS, Dos Santos EB, Lucas LDSM, et al.: Prevalence of neurological complications associated with Zika virus in a brazilian metropolis. Neurol Int. 2018; 10(2): 7638. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDel Carpio-Orantes L, Peniche Moguel KG, Sanchez Diaz JS, et al.: Guillain-Barré syndrome associated with Zika virus infection: Analysis of a cohort from the region of northern Veracruz in 2016-2017. Neurologia. 2018; pii: S0213-4853(18)30173-7. PubMed Abstract \| Publisher Full Text\n\nIkejezie J, Shapiro CN, Kim J, et al.: Zika Virus Transmission - Region of the Americas, May 15, 2015-December 15, 2016. MMWR Morb Mortal Wkly Rep. 2017; 66(12): 329–34. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMier-Y-Teran-Romero L, Delorey MJ, Sejvar JJ, et al.: Guillain-Barré syndrome risk among individuals infected with Zika virus: a multi-country assessment. BMC Med. 2018; 16(1): 67. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVieira MADCES, Cruz ACR, Barros ANM, et al.: Guillain-Barré syndrome and dengue-like disease in 2015: temporal relationship in Piauí state and implications on Zika virus surveillance. Rev Inst Med Trop Sao Paulo. 2017; 59: e22. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMalta JM, Vargas A, Leite PL, et al.: Guillain-Barré syndrome and other neurological manifestations possibly related to Zika virus infection in municipalities from Bahia, Brazil, 2015. Epidemiol Serv Saude. 2017; 26(1): 9–18. PubMed Abstract \| Publisher Full Text\n\nTolosa N, Tinker SC, Pacheco O, et al.: Zika Virus Disease in Children in Colombia, August 2015 to May 2016. Paediatr Perinat Epidemiol. 2017; 31(6): 537–45. PubMed Abstract \| Publisher Full Text\n\nBrenciaglia M, Noel TP, Fields PJ, et al.: Clinical, Serological, and Molecular Observations from a Case Series Study during the Asian Lineage Zika Virus Outbreak in Grenada during 2016. Can J Infect Dis Med Microbiol. 2018; 2018: 4635647. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVieira MADCES, Costa CHN, Linhares ADC, et al.: Potential role of dengue virus, chikungunya virus and Zika virus in neurological diseases. Mem Inst Oswaldo Cruz. 2018; 113(11): e170538. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nde Araujo TVB, Rodrigues LC, de Alencar Ximenes RA, et al.: Association between Zika virus infection and microcephaly in Brazil, January to May, 2016: preliminary report of a case-control study. Lancet Infect Dis. 2016; 16(12): 1356–63. PubMed Abstract \| Publisher Full Text\n\nBarreras P, Pamies D, Kumar A, et al.: 2016 Annual Meetings. Ann Neurol. 2016; 80(s20): S1–S432. PubMed Abstract \| Publisher Full Text\n\nPomar L, Malinger G, Benoist G, et al.: Association between Zika virus and fetopathy: a prospective cohort study in French Guiana. Ultrasound Obstet Gynecol. 2017; 49(6): 729–36. PubMed Abstract \| Publisher Full Text\n\nBrasil P, Pereira JP Jr, Moreira ME, et al.: Zika Virus Infection in Pregnant Women in Rio de Janeiro. N Engl J Med. 2016; 375(24): 2321–34. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHonein MA, Dawson AL, Petersen EE, et al.: Birth Defects Among Fetuses and Infants of US Women With Evidence of Possible Zika Virus Infection During Pregnancy. JAMA. 2017; 317(1): 59–68. PubMed Abstract \| Publisher Full Text\n\nCosta Monteiro LM, Cruz GNO, Fontes JM, et al.: Neurogenic bladder findings in patients with Congenital Zika Syndrome: A novel condition. PLoS One. 2018; 13(3): e0193514. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCao-Lormeau VM, Blake A, Mons S, et al.: Guillain-Barré Syndrome outbreak associated with Zika virus infection in French Polynesia: a case-control study. Lancet. 2016; 387(10027): 1531–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVandenbroucke JP: Case reports in an evidence-based world. J R Soc Med. 1999; 92(4): 159–63. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKopec JA, Esdaile JM: Bias in case-control studies. A review. J Epidemiol Community Health. 1990; 44(3): 179–86. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLipton P: Inference to the best explanation. Routledge; 2003. Reference Source\n\nMoore CA, Staples JE, Dobyns WB, et al.: Characterizing the Pattern of Anomalies in Congenital Zika Syndrome for Pediatric Clinicians. JAMA Pediatr. 2017; 171(3): 288–95. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGrubaugh ND, Ishtiaq F, Setoh YX, et al.: Misperceived Risks of Zika-related Microcephaly in India. Trends Microbiol. 2019; 27(5): 381–383. PubMed Abstract \| Publisher Full Text\n\nNutt C, Adams P: Zika in Africa-the invisible epidemic? Lancet. 2017; 389(10079): 1595–6. PubMed Abstract \| Publisher Full Text" }	[ { "id": "53510", "date": "18 Oct 2019", "name": "Rajesh Verma", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI think this is an excellent review written on the topic Zika virus and adverse fetal outcome and neurological complication (Guillain Barre Syndrome).\nThe studies interpretation in terms of statistical significance is outstanding.\nThe pathogenesis of these complications are not mentioned. How GBS develops after Zika infection? Whether immune mediated or virus has direct affinity for nervous system?\nWhat precautions to be taken at community level to curb the menace of Zika virus infection?\n\nIs the living method justified? Yes\n\nHave the search and update schedule been clearly defined and justified? Yes\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes\n\nIs the statistical analysis and its interpretation appropriate? Yes\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes", "responses": [] }, { "id": "55660", "date": "12 Nov 2019", "name": "Ana B.G. Veiga", "expertise": [ "Reviewer Expertise Molecular epidemiology", "arboviruses", "respiratory viruses" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article is a well-written systematic review of studies that show that Zika virus infection is associated with congenital abnormalities, and with Guillain-Barré syndrome. Because it is a living systematic review, the study gives interesting and important updates.\nThere are a few grammar errors that must be corrected:\nIn Methods, correct is \"This review and subsequent updates will focus on four...\" In the sentence \"Araujo et al. found a 73.1 (95% CI 13·0–Inf) times higher odds was reported for microcephaly...\", I would delete \"was reported\", so it would be better if written \"Araujo et al. found a 73.1 (95% CI 13·0–Inf) times higher odds for microcephaly when ZIKV infection was assessed by reverse transcription polymerase chain reaction (RT-PCR) in the neonate.\" (page 6) PRNT (plaque reduction neutralisation test) should be defined when first mentioned. Correct non traveller to non-traveller (pages 9 and 10).\nThe text mentions Figure 4 as showing the combined adverse congenital outcomes, and Figure 5 as showing microcephaly as outcome. However, this is not clear in the figures. In addition, why are the diagnostic tests considered outcomes in Figure 5?\n\nFigures should be self-explanatory, so I suggest mentioning GBS in the legend of Figure 6 so that the reader doesn't need to refer to the text. Also, VNT should be specified as virus neutralisation test in the figure legend.\n\nIs the living method justified? Yes\n\nHave the search and update schedule been clearly defined and justified? Yes\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes\n\nIs the statistical analysis and its interpretation appropriate? Yes\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1433
https://f1000research.com/articles/8-1431/v1	14 Aug 19	{ "type": "Brief Report", "title": "Ethanol as a potential mosquito sample storage medium for RNA preservation", "authors": [ "Mirsha G. Torres", "Allison M. Weakley", "James D. Hibbert", "Oscar D. Kirstein", "Gregory C. Lanzaro", "Yoosook Lee", "Mirsha G. Torres", "Allison M. Weakley", "James D. Hibbert", "Oscar D. Kirstein", "Gregory C. Lanzaro" ], "abstract": "Sample storage for downstream RNA analysis can be challenging in some field settings, especially where access to cryogenic materials or refrigeration/freezer facilities are limited. This has limited RNA-based studies on African malaria vectors collected in the field. We evaluated RNA quality after storing mosquito samples in three different sample preservation media over a 4-week period. Storing mosquito specimens in cold (4°C) media significantly improved yields of intact RNA. Our results indicate commercially available products perform well in keeping RNA integrity as advertised. Moreover, absolute ethanol may be an economical alternative for sample preservation that can be utilized in some resource-limited settings.", "keywords": [ "Sample storage", "RNA preservation", "field settings", "mosquito", "genetic analysis" ], "content": "Introduction\n\nSamples to be used for downstream RNA analysis (e.g. RNA-Seq) are typically preserved by snap-freezing using liquid nitrogen or dry ice and then stored at -80°C until RNA extraction1,2. Several protocols have been published for preservation and extraction of genetic material from field collected samples3,4. Along with these protocols, there are products available to preserve nucleic acids from field collected specimens. These products include Allprotect Tissue Reagent (Qiagen, Hilden, Germany) and RNAlater (Thermo Fisher Scientific, Waltham, MA, USA). These reagents can stabilize tissue samples to maintain RNA content for one (RNALater) to six months (AllProtect) at mildly cold (4°C) temperatures. The duration can be increased to over one year if samples are stored in colder (-20°C) temperatures.\n\nOptimal preservation of field collected samples to be used for gene expression studies require high quality nucleic acid, requiring preservation and stabilization of the RNA molecule5. Unfortunately, the preservation of genetic material for expression studies based on field samples is difficult, and cryopreservation is often not possible. This is particularly applicable to field collections of Anopheles mosquitoes, which are the prime vector of malaria parasites6 and exist in remote areas of Africa.\n\n\nMethods\n\nA total of 54 laboratory-reared Anopheles coluzzii mosquitoes from the UC Davis Vector Genetics Laboratory insectarium were subjected to various sample preservation conditions, as listed in Table 1. Three different sample preservation solutions were tested: Allprotect Tissue Reagent (Qiagen, Hilden, Germany), RNAlater (Thermo Fisher Scientific, Waltham, MA, USA), and 100% ethanol. Each set of samples was maintained in one of the three preservation solutions and subjected to two different temperature settings: typical refrigeration temperature (4°C) or at room temperature (28°C). A total of nine mosquito samples were stored in each of the six conditions listed in Table 1 for 4 weeks prior to RNA extraction.\n\nFollowing a four-week sample preservation period, RNA was extracted from each mosquito sample using the Qiagen AllPrep DNA/RNA Mini Kit (Qiagen, Hilden, Germany) employing the manufacturer's protocol. The RNA concentration was measured using a Qubit RNA High Sensitivity kit and Qubit 2.0 instrument (Thermo Fisher Scientific, Waltham, MA, USA) using the manufacturer’s protocol. The RNA fragment size distribution was examined using the Agilent High Sensitivity RNA Analysis kit and TapeStation 4200 instrument (Agilent, Santa Clara, CA, USA), and the dominant peak size and proportion of long (>1000 bp) fragments were recorded. Typical RNA integrity number (RIN) which measures the 28S and 18S rRNA ratio was not used due to negligible 28S peaks, which is typical for insect RNA extracts7.\n\nMann-Whitney tests were conducted using the scipy module version 1.2.08 in the Jupyter notebook9 version 4.1. environment. Plots were generated using Matplotlib version 3.1.010.\n\n\nResults and Discussion\n\nResults for each sample are available as Underlying data11. Preservation conditions that resulted in the highest concentration of longer RNA fragments (>1000 bp) were considered to be best for downstream genetic analysis, as opposed to those resulting in degraded RNA (fragment size < 1000 bp). As expected, storage at 4°C generally preserved RNA integrity better than 28°C. There were no significant differences in total RNA concentration or dominant peak size between samples stored in AllProtect™ or RNALater™ at either storage temperature (Mann-Whitney test, P>0.05, Figure 1). Samples stored in absolute ethanol, however, showed a significant increase in RNA yield (Mann Whitney Test, P=0.0065) and significant decrease in dominant peak size (Mann-Whitney test, P=0.00020) when stored at 28°C. We observed a significant reduction in long fragment (>1000 bp) RNA in samples stored at 28°C than at 4°C regardless of the preservation solution (Mann-Whitney test, ɑ < 0.05). These results suggest, as expected, that higher temperatures accelerates tRNA degradation. Degradation was greater in absolute ethanol, decreasing the proportion of long (>1000bp) fragments from 74.1% (±5.6 STD) at 4°C to 16.9% (±10.1 STD) at 28°C. RNAlater and AllProtect™ maintained a >40% content of RNA fragments of ≥1000 bp.\n\nAt 4°C no significant difference was observed in dominant RNA peak size (1600-1814 bp) and proportion of long (>1000bp) RNA fragments (62.2–74.9%) among the three preservation media tested (Mann-Whitney test, P>0.05). The only significant difference was a lower concentration of RNA in ethanol compared with the other preservation solutions (Mann-Whitney test, P<0.0081).\n\nAbsolute ethanol did not preserve RNA integrity at 28°C, with only 16.9% (±10.1 STD) of RNA content composed of 1000 bp or longer fragments. However, the peak RNA fragment size for samples stored in absolute ethanol at 28°C was 869 bp (±73 STD) showing little variation (Figure 1) yet similar concentrations as the other two preservation solutions. This quality may be sufficient to conduct downstream RNA analysis for real-time PCR or RNA-Seq.\n\nOverall, samples stored in RNAlater™ or AllProtect™ at 4°C provide satisfactory preservation of RNA content from field collected samples after 4 weeks in storage. Absolute ethanol may provide an economical alternative in resource-constrained field settings. Currently in the USA, AllProtect™ is available at ~$6.5/mL, RNAlater™ at $0.9-1.4/mL, and 200 proof lab grade ethanol at $0.1-0.6/mL. When stored at 4°C, absolute ethanol may be a viable alternative to commercially available products. Although RNA stored in ethanol at 28°C will degrade faster, it nonetheless maintained fragment sizes over 800 bp after 4 weeks in storage. Future evaluation of RNA quality utilizing real time PCR or RNA-seq may be necessary to elucidate whether ethanol is indeed an adequate sample preservation solution for RNA preservation. For practical applications, keeping specimens in a commercial RNA storage solution at 4°C maximizes maintenance of RNA integrity.\n\n\nData availability\n\nOpen Science Framework: Sample storage condition testing for RNA preservation. https://doi.org/10.17605/OSF.IO/BRNPV11.\n\nThis project contains data on RNA source, storage, concentration, dominant peak size and quality from each sample assessed in this study.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).", "appendix": "Grant information\n\nThis work was supported by the University of California Irvine Malaria Initiative.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgement\n\nWe thank Ms. Kendra Person (UC Davis) for providing samples used in this study.\n\n\nReferences\n\nJuneja P, Ariani CV, Ho YS, et al.: Exome and transcriptome sequencing of Aedes aegypti identifies a locus that confers resistance to Brugia malayi and alters the immune response. PLoS Pathog. 2015; 11(3): e1004765. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhao L, Alto BW, Shin D: Transcriptional Profile of Aedes aegypti Leucine-Rich Repeat Proteins in Response to Zika and Chikungunya Viruses. Int J Mol Sci. 2019; 20(3): pii: E615. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDrakulovski P, Locatelli S, Butel C, et al.: Use of RNAlater as a preservation method for parasitic coprology studies in wild-living chimpanzees. Exp Parasitol. 2013; 135(2): 257–61. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChoo JM, Leong LE, Rogers GB: Sample storage conditions significantly influence faecal microbiome profiles. Sci Rep. 2015; 5: 16350. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCamacho-Sanchez M, Burraco P, Gomez-Mestre I, et al.: Preservation of RNA and DNA from mammal samples under field conditions. Mol Ecol Resour. 2013; 13(4): 663–73. PubMed Abstract \| Publisher Full Text\n\nManguin S: ANOPHELES MOSQUITOES - NEW INSIGHTS INTO MALARIA VECTORS Preface. Anopheles Mosquitoes - New Insights into Malaria Vectors. 2013; Xii-Xiii. Publisher Full Text\n\nWinnebeck EC, Millar CD, Warman GR: Why does insect RNA look degraded? J Insect Sci. 2010; 10(1): 159. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJones E, Oliphant E, Peterson P, et al.: SciPy: Open Source Scientific Tools for Python. 2001. Reference Source\n\nJupyter Team: Project Jupyter. 2015. Reference Source\n\nHunter JD: Matplotlib: A 2D graphics environment. Comput Sci Eng. 2007; 9(3): 90–5. Publisher Full Text\n\nLee Y: Sample storage condition testing for RNA preservation. 2019. http://www.doi.org/10.17605/OSF.IO/BRNPV" }	[ { "id": "52797", "date": "27 Aug 2019", "name": "Tara Roth", "expertise": [ "Reviewer Expertise Infectious disease", "real-time PCR", "disease ecology", "vector borne disease" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a useful methods paper to help clarify any misconceptions researchers might have about what is strictly necessary in order to preserve genomic specimens. I feel like this information could be highly useful for field researchers but also as information to help students just starting research projects.\nOverall, I don't have many comments to make. The sample size is somewhat small but appears statistically valid. The one thing that would have been really nice to see is a comparison of the study findings to RNA degradation under ideal conditions (-80) as well as with no preservation methods - just dry on the table.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "52465", "date": "27 Aug 2019", "name": "Megan A. Riddin", "expertise": [ "Reviewer Expertise My research area is Medical Entomology", "with focus on malaria vectors in southern Africa. My research pertains to mosquito vector epidemiology and control." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article addresses the efficacy of commercially available RNA preservation media and absolute ethanol at two temperatures, room temperature (28°C) and cold (4°C). The article focuses upon a critical topic that is the success of preservation of samples for downstream RNA analysis, particularly to find a reliable source in limited field source settings. The outcome of this study has direct application to a number of fields, including the study of malaria vectors which are often collected in preservation limited settings with a complete lack of freezer facilities for long periods.\nThis paper is a significant addition to the literature and has been addressed well. The selection of only a 4 week testing period is short, however, I feel the outcomes of this study can be immediately applied in a number of research fields, particularly research on mosquito disease vectors, as most field collections do not extend past such a period. By including absolute ethanol and a two RNA-preservative media options, this enables the covering of a cheaper and readily available option as well as other more stable, but more expensive and may be difficult to obtain, options for regions where snap-freezing is not an option.\nMore specific comments:\nThe authors covered the problem, objective and study highlights well in the abstract and state that absolute ethanol is a viable option in resource-limited settings.\nI can highly appreciate that this was a direct study for the preservation of malaria vectors in the field which is predominantly in resource limited regions. It would complement the methods to include reasoning on selection of insectary-reared Anopheles coluzzii, 9 individuals and the 4 week period. This would be of interest to researchers who would replicate the methodology or adapt it for their sample and preservations needs. I commend the authors on including RNA concentration as well as fragment size distribution analysis, offering a comprehensive effect of the preservation method on RNA integrity.\nThe results of the study are clearly and concisely described, and significance provided. The most important results are communicated and basis for conclusions are of high standard. The figures are understandable and visually descriptive. It is excellent that the authors further include an expense comparison for the appreciation of the use of absolute ethanol in funding and resource limited studies, as well as highlight the need for further evaluation of the preservation methods for real time PCR and RNA-seq.\nOverall this is a worthy article which is a good addition to literature, and has direct application for RNA preservation.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? I cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1431
https://f1000research.com/articles/8-1424/v1	13 Aug 19	{ "type": "Review", "title": "Role of microsatellites in genetic analysis of Bombyx mori silkworm: a review", "authors": [ "Julian David Trochez-Solarte", "Ximena Ruiz-Erazo", "Martha Almanza-Pinzon", "Giselle Zambrano-Gonzalez", "Ximena Ruiz-Erazo", "Martha Almanza-Pinzon", "Giselle Zambrano-Gonzalez" ], "abstract": "In the genome of Bombyx mori Linnaeus (1758), the microsatellites, or simple sequence repeats (SSR), feature among their particular characteristics a high adenine and thymine (A/T) content, low number of repeats, low frequency, and a grouping in \"families\" with similar flanking regions. Such characteristics may be the result of a complex interaction between factors that limit the size and dispersion of SSR loci—such as their high association with transposons—and mean that microsatellites within this taxon suitable as molecular markers are relatively rare. The determination of genetic profiles in populations and cell lines has not been affected owing to the high level of polymorphism, nor has the analysis of diversity, structure and genetic relationships. However, the scarcity of suitable microsatellites has restricted their application in genetic mapping, limiting them to preliminary identification of gene location of genes or quantitative trait loci (QTLs) related to thermotolerance, resistance to viruses, pigmentation patterns, body development and the weight of the cocoon, the cortex, the pupa and the filament. The review confirms that, as markers, microsatellites are versatile and perform well. They could thus be useful both to advance research in emerging countries with few resources seeking to promote sericulture in their territories, and to advance in the genetic and molecular knowledge of characteristics of productive and biological interest, given the latest technological developments in terms of the sequencing, identification, isolation and genotyping of SSR loci.", "keywords": [ "Bombyx mori", "silkworm", "molecular marker", "sericulture", "Simple Sequence Repeats", "SSR." ], "content": "Introduction\n\nDomesticated around 5,000 years ago, the silkworm, Bombyx mori L. is the basis of sericulture, an agroindustrial activity that involves the breeding of silkworms, cultivation of Morus spp. mulberry plants as the sole source of food, and industrial processing of cocoons to produce natural silk yarn1. B mori is an ideal organism as an experimental animal for genetic and biological research. Silkworms are easily reared and produce a genetically uniform population. This, added to their economic importance, has made B. mori one of the most widely studied insects2,3. As a result, genetic materials of agronomic and scientific interest have been identified, characterized and conserved4 using a variety of molecular markers, prominent among them microsatellites, or simple sequence repeats (SSRs).\n\nMicrosatellites are regions of DNA in which a sequence or motif between 1 and 6 bp is repeated in tandem 5 to 100 times, the number of repeats of the same locus being highly variable both inter- and intra-populationally. They are also ubiquitous, distributed uniformly in eukaryotic genomes5. SSR loci are usually flanked by regions of unique sequence, which allows PCR amplification with specifically designed primers and determination of the specific genetic profile or genotype of an individual for several loci through the pattern of bands displayed on capillary sequencing equipment5. These markers are codominant and highly polymorphic. They have greater reproducibility and band comparability, are less sensitive to contamination by foreign DNA (due to their specificity) and can be amplified using fragmented or partially degraded DNA, because of the reduced size of the loci6.\n\nThese molecular markers have been a useful tool in sericulture, especially in the management, characterization and conservation of materials in germplasm banks7,8 and in the development of genetically improved materials9. They have been particularly difficult to apply, however, to genetic mapping, mainly due to the low frequency of loci suitable as molecular markers in the B. mori genome. Their contribution in this area has therefore been restricted to preliminary mapping of some genes and QTLs related to characteristics of productive interest10 or to processes of B. mori body development and pigmentation11,12.\n\nThe review focuses on the description and analysis of the specific characteristics of microsatellites in the B. mori genome, the role these repetitive DNA regions have played as molecular markers in this organism, and how these would play an important role in future in the genetic analysis, conservation and use of silkworm germplasm.\n\n\nCharacteristics of silkworm genome microsatellites\n\nMicrosatellite characteristics—distances between loci, abundance, distribution, motif and the average number of repeats they comprise—may vary among taxa13. The B. mori microsatellites feature average distances of between 49 and 161 kb14,15, their genome coverage is only 0.31%16 and they have a cloning efficiency of between 0.77 and 3.5%14,15. Together, such values indicate that microsatellites in the B. mori genome are rare. The pattern is not unusual however, having been observed in other insect species, including several in the Drosophila genus17,18.\n\nSSR loci with mononucleotide-like motifs represent 60% of the microsatellites of the B. mori genome, high compared to other insect species17, and are formed by a low number of repeats, nine on average among the different repeat motifs. Loci of more than 15 repeats are scarce, except for mononucleotide motifs16,17. Harr and Schlötterer19 report that Drosophila melanogaster Meigen (1830) also has short SSR loci on average (<15 repeats), because the greater the length, the greater the tendency of the D. melanogaster microsatellites to mutations that decrease the number of repeats instead of increasing it. This is speculated to be the result of interaction between factors related to DNA repair and replication mechanisms.\n\nThe B. mori genome has a comparatively high composition of regions rich in adenine and thymine (A/T). Moreover, the microsatellites that include a high proportion of these bases are the most abundant, constituting approximately 40% of the dinucleotide motifs and 28.3% of the total20. Such a predominance has not been seen in other organisms16,20.\n\nTransposons appear to play a key role in the evolution of microsatellites in the B. mori genome21. Of the B. mori SSR loci, 35% are grouped into \"families\" with flanking regions similar in sequence. This is due to their association with transposons17, mobile genetic elements capable of replicating themselves and associated regions. Transposons account for 35–43.6 % of the B. mori genome22,23. For Meglécz et al.17, this high association suggests one of two scenarios: transposons might favor the development of microsatellites in their vicinity, or promote their formation at the moment of transposition. It is possible, however, that they have no part to play in the genesis. The SSR loci could develop independently, but have a structure that favors the insertion of mobile elements24.\n\nThe characteristics of the B. mori microsatellites—low frequency, low average number of repeats, and grouping in families—are common to the Lepidoptera order. They explain the difficulties in isolating single copy microsatellites17,18,25,26, but the advance in high-throughput sequencing techniques and graph-based cluster analysis27, as well as screening against transposons elements in the isolation procedure28 may in future facilitate identifying suitable microsatellites for genetic studies in species in this taxon.\n\nZhang21 suggested that the characteristics of Lepidoptera microsatellites imply that, in their respective genomes, most have experienced a recent development and multiplication, in the different lepidopteran species. This may mirror that reported in D. melanogaster19, characterized because its long SSR loci are of recent origin and have short prevalence periods.\n\nMicrosatellite evolution is extremely complex24. The characteristics in B. mori and the Lepidoptera, as well as in other organisms29, appear to result from interaction in the genome between the mutations or events that cause them and the factors that obstruct or limit their development, with a balance in favor of the latter in silkworm. The key to the low average number of repeats and low frequency of SSR loci in the B. mori genome may be the abundance of A and T bases, and of microsatellites composed of these bases. Regions rich in A/T have a higher frequency of double-strand breaks in the DNA, which can induce and facilitate both the loss of nucleotides during non-homologous recombination30 and the insertion of transposons able to divide an SSR locus in two and interrupt its development24,31,32. This could explain why A/T-rich microsatellites in B. mori have a lower average number of repeats, as reported by Zhan et al.20.\n\nAdditionally, the presence in B. mori of an efficient DNA mismatch repair mechanism, the system in charge of correcting the incorrect incorporation of bases29, could counteract replication slippage, the major mutational mechanism in explaining the origin and evolution of repetitive DNA regions24.\n\n\nApplications of microsatellites\n\nMicrosatellites have proved a useful tool for generating band patterns that enable discrimination of silkworm lines. Kim et al.8,33,34 found that 25–28% of the amplified alleles are specific to the lines, due to which a small number of microsatellites, one to three, allow identifying approximately 20% of the analyzed materials without resorting to the genotyping of other loci. Hou et al.35 and Chandrakanth et al.36 likewise indicate that the analyzed genotypes are homozygous for a substantial part of the microsatellites used. Since each locus is multiallelic, those that amplify as a single band in each line are powerful tools for identification, as demonstrated by Li et al.37 when discriminating between two closely related lines identical in their morphological characteristics: Dazao and P50.\n\nDNA fingerprinting also represents a tool for the identification of insect cell lines. Between 3 and 8 microsatellites have therefore been used to generate profiles of cell lines susceptible to the nuclear polyhedrosis virus of B. mori (BmNPV), developed to study replication and expression mechanisms of the virus38–40. McIntosh et al.41 obtained stable DNA profiles even after performing 200 subcultures by using coding regions (aldolase, prolactin receptor, interleukin-1β) as molecular markers. Microsatellites are less stable, however, due to their high mutation rate—between 10-4 and 10-6 per locus per generation42. Thus, for future identification of cell lines, it is advisable to evaluate and select SSR loci that present relatively low mutation rates.\n\nSericulture depends on the strategic use of silkworm germplasm to develop hybrids with high yields of silk that resist or tolerate disease and adverse climatic conditions43, based on knowledge of the extent and distribution of the genetic diversity available in both the domesticated silkworm B. mori and its wild relative Bombyx mandarina Moore (1872). In this context, between 500 and 700 microsatellites were developed in B. mori, of which 5 to 27 markers have been used for analyzing genetic diversity (Table 1).\n\n† SSRs: microsatellite loci used (Simple Sequence Repeats)\n\n‡ PIC: polymorphic information content (PIC) ranking of microsatellites within the study\n\nMiao et al.15 and Zhan et al.20 discovered that the genome of B. mori lines with contrasting characteristics is similar, in terms of the low percentage of polymorphic SSR loci, ranging between 17% and 24% compared, for example, with 85% for the European bee, Apis mellifera44, and 55% in laboratory rats45. Together, these data attribute the origin of B. mori to a single domestication event from a reduced population of B. mandarina15. Xia et al.46, however, on conducting a complete genome analysis on domesticated lines and wild individuals report that B. mori harbors 83% of the genetic variability of wild populations, indicating that the origin was possibly not limited to a reduced population or a single domestication event43.\n\nThe low percentage of polymorphic SSR loci found among domesticated lines of B. mori should not be interpreted as evidence of the reduction of genetic diversity with respect to wild populations. The low figure could instead be the result of size homoplasy, a process of change by which convergent mutations cause microsatellites, belonging to different lineages, to have the same length in base pairs47.\n\nSize homoplasy is favored when mechanisms are present that neutralize elongation of the microsatellites and limit the number of repeats, since possible alleles are reduced and mutations are more likely to converge in the same length. These conditions appear to be present in the B. mori genome, in which most microsatellites are of reduced size16,17,20. In 2016, De Barba et al.48 proposed a new method for genotyping microsatellites using high-throughput sequencing; this would allow direct access to the microsatellite sequences in B. mori and evaluate whether the low percentage of polymorphic loci is due to the presence of size homoplasy.\n\nAlthough B. mori has not experienced a drastic reduction in genetic diversity compared to B. mandarina, the wide genetic distances between domesticated and wild populations49 indicate that these species have a marked genetic differentiation and distant relationships. This is likely due to the absence of genetic flow, a result of the inability of B. mori to fly and to survive without human intervention50. Thus, B. mandarina represents a potential unique source of genetic material for sericulture43.\n\nCluster analyzes to determine the relationships between B. mori lines provide contradictory results. Reddy et al.14, Qian et al.51, Thiyagu and Kamble52 and Chandrakanth et al.36 report that grouping the materials based on the microsatellites analyzed corresponded to type of voltinism, geographical origin, silk productivity, color or shape of the cocoon. However, the groups formed in studies that analyzed a larger sample of germplasm—69 lines on average, compared to 18 in the works cited above—exhibit mixtures of genotypes with variability in these characteristics8,33–35,37.\n\nSeveral different scenarios may explain why B. mori genotypes, with apparently diverse traits, are grouped together, the main one being the hybridization that has historically been used, in pure lines, to perform the introgression of genes that increase silk yield or survival rate in various environmental conditions55–58. This would alter their phenotypic characteristics but, due to backcrossing, they would maintain a similar—or even the same—genotype, depending on the microsatellites analyzed.\n\nVarious pure lines may also share an origin in the same ancestral population of B. mandarina, but would have been selected to express different phenotypes, maintaining a high similarity at the genotypic level7,37,59. This might also explain why pure lines with similar characteristics are not always grouped, given that they would have a distant relationship, but would have been selected to express similar traits35,37.\n\nTraditional genetic improvement in B. mori implies the use of phenotypic characteristics or geographical origin to differentiate and select parents with contrasting characteristics with which to perform crosses. However, due to the scenarios already mentioned, these characteristics do not always make it possible to accurately determine genetic relations between materials60. Microsatellites thus represent an important tool for making accurate estimates of genetic diversity and relationships in order to develop genetically improved materials (hybrids), bearing in mind that, comparatively, performance is better than that of other markers such as RAPDs, RFPLs and ISSRs6.\n\nMicrosatellites not only represent a marker with a good performance in analysis of genetic diversity in B. mori. They constitute a versatile tool that can be adjusted to meet the research requirements and the resources available. For example, they can potentially be genotyped with high-throughput sequencing for greater accuracy48 and even be identified and analyzed simultaneously with SNPs to strengthen inferences about diversity, structure and genetic relationships61. However, they can also be identified by means of polyacrylamide gels or capillary sequencing equipment when fluorescently labeled5. As such, they provide accessible and profitable information for managing germplasm banks and promoting regional initiatives in developing countries that do not have the resources to access cutting-edge technology, yet view sericulture as an opportunity to generate employment and improve the conditions of the rural population1.\n\nDetermination of the relative positions of microsatellite markers in the chromosomes of the B. mori genome began with the low-density linkage map developed by Prasad et al.16. The medium density one constructed by Miao et al.15 followed, with an average distance between markers of 6.3 cM and 29 linkage groups. Zhan et al.20 subsequently increased the density of this map using new lines and mapping populations of B. mori, decreasing the average distance between markers to 4.8 cM (Table 2).\n\n†LGs: linkage groups\n\n‡SSRs: microsatellite loci (Simple Sequence Repeats)\n\n§Numbers in square brackets indicate linkage groups assigned to established linkage groups\n\n¶BF1: backcross\n\n††BC1M: backcross with F1 male\n\nThe density achieved with the linkage map of SSR markers is below the results expected by the authors due to the high homology (low percentage of polymorphic loci) between the loci of the B. mori lines used to generate the mapping populations15,20. Nevertheless, the resolution is sufficient to carry out the preliminary gene screening (Table 3) and identification of QTLs.\n\n†LG: linkage group where the gene is located\n\nExclusive linkage maps for the Z chromosome were developed by Nagaraja et al.62 and Miao et al.63 with the purpose of contributing to identification of genes related to control of the duration of larval stages, diapause, moltinism, body size and color, etc., and to analyze the role played by characteristics linked to sex in evolutionary processes in Lepidoptera, and differentiation of geographic races64.\n\nIdentification of SSR markers linked to genes was carried out in order to understand the molecular basis of characteristics of agronomic and scientific interest. The SSR loci identified by Miao et al.15 and Zhan et al.20 enabled the identification of genes related to characteristics such as thermotolerance, resistance to the Z strain of the virus of densonosis in B. mori, tolerance to fluorinated compounds, and absence of wing scales; studies that were used to develop improved lines with marker-assisted selection9,65,66. Genes related to pigmentation patterns were also identified, in cocoons and larvae, and in development processes such as the formation of extremities, thoracic segments, cell adhesion and regulation of molting (Table 3).\n\nMicrosatellites have also allowed tracking of QTLs in B. mori related to weight of cocoon, cortex, pupa and filament10,20, which are mainly located on chromosome 1 where they are strongly linked to SSR loci (LOD>11.0), contribute significantly to phenotypic variation (30 %) and have a simultaneous effect on the above characteristics. The location of the QTLs in chromosome 1 was delimited to a region of 290 kb, in which 12 candidate genes were identified that will allow study of the molecular mechanisms underlying these characters of agronomic interest in B. mori10. Additionally, Gao et al.67 discovered that Bombyx mori Nuclear Polyhedrosis Virus (BmNPV) resistance is a polygenic characteristic.\n\nLinkage maps with microsatellites have allowed us to advance in the knowledge of characteristics of economic interest in B. mori, as well as in the Lepidoptera genetic architecture. However, Xu et al.68 and Li et al.69 indicate that the most used models have not been adjusted according to the particular genetic characteristics of the silkworm, because they ignore the effects of gender and the presence of aquiasmatic meiosis, a process that implies the absence of chromosomal cross-linking in the germinal line of the females. As a result, these authors have proposed models that allow correcting the potential biases and lack of precision of the most used methods for mapping QTLs in B. mori. Xu et al.68 propose a statistical model to analyze QTLs in F2 populations, while Li et al.69 focus their method on the use of backcrossing and the rational selection of mapping populations to first identify the chromosomes with QTLs and subsequently their position.\n\n\nConclusions and future work\n\nThe microsatellites in B. mori have particular characteristics such as low frequency, low average number of locus repeats and grouping into \"families\". These are shared by Lepidoptera and seem to indicate that factors or mechanisms exist within the genome of this taxon that limit growth and stability of these repetitive DNA regions and their used as single copy loci molecular markers. Such as an abundance of loci SSR rich in A/T susceptible to double strand breaks and loss of repeats, as well as the possible existence of efficient DNA repair mechanisms that avoid the incorrect incorporation of bases (DNA mismatch repair) and the high association with transposons that would cause the formation of groups of loci with high similarity in their flanking regions.\n\nThe characteristics of the SSR loci in B. mori, and generally in all Lepidoptera, have made identification and isolation of these markers difficult. They have also been a limitation for applications in genetic mapping. The low frequency and high homology (low percentage of polymorphic loci) of the microsatellites between contrasting lines of B. mori, possibly due to the size homoplasy, has not allowed the development of high-density linkage maps. This, together with the absence of mapping models adjusted to this organism, has hindered identification of genes and QTLs, limiting contributions mainly to preliminary mapping, for example of regions related to pigmentation patterns and development processes, as well as to weight of cocoon, cortex, pupa and filament.\n\nThese regions of repetitive DNA, however, have shown a high discriminating power between B. mori lines due to the high level of polymorphism, the finding of a percentage of single alleles higher than 20%, and the high levels of homozygosity in the materials analyzed, so that a reduced subset of 5–8 SSR loci have made it possible to generate DNA fingerprints, estimating the genetic diversity of domesticated and wild materials and determining the genetic relationships between closely related lines such as Dazao and P50. Although these markers additionally represent a potential tool for identifying B. mori cell lines, it is necessary to evaluate and select a subset of microsatellites with relatively low mutation rates that provide stability to the genetic profiles for 200 or more subcultures.\n\nThe data indicate that microsatellites will continue to be important for the study, management, conservation and use of silkworm germplasm. They have shown superior performance in these aspects compared to most molecular markers and are versatile – they can be analyzed, depending on resources available and expected reliability, with traditional polyacrylamide gels, with analysis of DNA fragments marked with fluorescence, or with the latest sequencing technologies. The latter, in addition to providing greater precision and automation in genotyping, would also facilitate identification of suitable SSR loci as molecular markers and allow simultaneous analysis with single nucleotide polymorphisms (SNPs), which would complement and strengthen the inferences and analyzes obtained when using them separately.\n\nIn this context, microsatellites would play an important role both in supporting the research carried out in B. mori germplasm banks in emerging countries wishing to promote sericulture in their territories, but that do not have the resources to access cutting-edge technologies and in advancing understanding of the complex genetic and molecular mechanisms underlying characteristics of productive and biological interest.\n\n\nData availability\n\nNo data are associated with this article.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgments\n\nThe authors are grateful to the University of Cauca and the following research groups for the scientific support: Integrated Systems of Agricultural, Forestry and Aquaculture Production (SISINPRO), Geology, Ecology and Conservation (GECO); we are also indebted to the \"Technological Development for the Obtaining of Organic and Innovative Products of Natural Silk\" project of the General System of Royalties and the Government of Cauca; and we are especially grateful to Colin McLachlan for suggestions related to the English text.\n\n\nReferences\n\nSohn KW: Technical manual for tropical sericulture: practical technology to produce silkworm eggs and cocoons in the tropics. Kigali, Republica de Ruanda: Korea International Cooperation Agency (KOICA). 2014.\n\nPetkov NI, Tzenov PI, Petkov ZM, et al.: Silkworm, Bombyx mori L. germplasm resources in Bulgaria. National Centre for Agrarian Sciences–Sofia, North-West Regional Agrotechpark - Vratza, Regional Centre for Scientific Applied Service - Vratza, Sericultural Experiment Station–Vratza. Sofia, Bulgaria: PublishScieSet – Eco. 2006; 264. Reference Source\n\nZhou Z, Yang H, Zhong B: From genome to proteome: great progress in the domesticated silkworm (Bombyx mori L.). Acta Biochim Biophys Sin (Shanghai). 2008; 40(7): 601–11. PubMed Abstract \| Publisher Full Text\n\nGoldsmith MR: Recent progress in silkworm genetics and genomics. In: Goldsmith MR, Frantisek M, editors. Molecular biology and genetics of the lepidoptera. New York: CRC press, 2009; 25–48. Reference Source\n\nMadesis P, Ganopoulos I, Tsaftaris A: Microsatellites: evolution and contribution. In: Kantartzi KS, editor. Microsatellites: methods and protocols. Totowa, NJ: Humana Press, Methods Mol Biol. 2013; 1006: 1–13. PubMed Abstract \| Publisher Full Text\n\nNagaraju J, Reddy KD, Nagaraja GM, et al.: Comparison of multilocus RFLPs and PCR-based marker systems for genetic analysis of the silkworm, Bombyx mori. Heredity (Edinb). 2001; 86(Pt 5): 588–97. PubMed Abstract \| Publisher Full Text\n\nFurdui EM, Mărghitaş LA, Dezmirean DS, et al.: Genetic characterization of Bombyx mori (Lepidoptera: Bombycidae) breeding and hybrid lines with different geographic origins. J Insect Sci. 2014; 14(1): pii: 211. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKim KY, Kang PD, Ryu KS, et al.: Microsatellite analysis of the silkworm strains (Bombyx mori) originated from China. Int J Ind Entomol. 2012; 25(1): 81–92. Publisher Full Text\n\nLi MW, Yu HJ, Yi XL, et al.: Marker-assisted selection in breeding silkworm strains with high tolerance to fluoride, scaleless wings, and high silk production. Genet Mol Res. 2015; 14(3): 11162–70. PubMed Abstract \| Publisher Full Text\n\nLi B, Wang XY, Hou CX, et al.: Genetic analysis of quantitative trait loci for cocoon and silk production quantity in Bombyx mori (Lepidoptera: Bombycidae). Eur J Entomol. 2013; 110(2): 205–13. Publisher Full Text\n\nDai F, Qiao L, Cao C, et al.: Aspartate Decarboxylase is Required for a Normal Pupa Pigmentation Pattern in the Silkworm, Bombyx mori. Sci Rep. 2015; 5: 10885. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWu F, Wang P, Zhao Q, et al.: Mutation of a Cuticle Protein Gene, BmCPG10, Is Responsible for Silkworm Non-Moulting in the 2nd Instar Mutant. PLoS One. 2016; 11(4): e0153549. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTóth G, Gáspári Z, Jurka J: Microsatellites in different eukaryotic genomes: survey and analysis. Genome Res. 2000; 10(7): 967–81. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nReddy KD, Abraham EG, Nagaraju J: Microsatellites in the silkworm, Bombyx mori: abundance, polymorphism, and strain characterization. Genome. 1999; 42(6): 1057–65. PubMed Abstract \| Publisher Full Text\n\nMiao XX, Xu SJ, Li MH, et al.: Simple sequence repeat-based consensus linkage map of Bombyx mori. Proc Natl Acad Sci U S A. 2005; 102(45): 16303–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPrasad MD, Muthulakshmi M, Madhu M, et al.: Survey and analysis of microsatellites in the silkworm, Bombyx mori: frequency, distribution, mutations, marker potential and their conservation in heterologous species. Genetics. 2005; 169(1): 197–214. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMeglécz E, Anderson SJ, Bourguet D, et al.: Microsatellite flanking region similarities among different loci within insect species. Insect Mol Biol. 2007; 16(2): 175–85. PubMed Abstract \| Publisher Full Text\n\nVan’t Hof AE, Brakefield PM, et al.: Evolutionary dynamics of multilocus microsatellite arrangements in the genome of the butterfly Bicyclus anynana, with implications for other Lepidoptera. Heredity (Edinb). 2007; 98(5): 320–8. PubMed Abstract \| Publisher Full Text\n\nHarr B, Schlötterer C: Long microsatellite alleles in Drosophila melanogaster have a downward mutation bias and short persistence times, which cause their genome-wide underrepresentation. Genetics. 2000; 155(3): 1213–1220. PubMed Abstract \| Free Full Text\n\nZhan S, Huang J, Guo Q, et al.: An integrated genetic linkage map for silkworms with three parental combinations and its application to the mapping of single genes and QTL. BMC Genomics. 2009; 10(1): 389. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhang DX: Lepidopteran microsatellite DNA: redundant but promising. Trends Ecol Evol. 2004; 19(10): 507–9. PubMed Abstract \| Publisher Full Text\n\nInternational Silkworm Genome Consortium: The genome of a lepidopteran model insect, the silkworm Bombyx mori. Insect Biochem Mol Biol. 2008; 38(12): 1036–45. PubMed Abstract \| Publisher Full Text\n\nOsanai-Futahashi M, Suetsugu Y, Mita K, et al.: Genome-wide screening and characterization of transposable elements and their distribution analysis in the silkworm, Bombyx mori. Insect Biochem Mol Biol. 2008; 38(12): 1046–57. PubMed Abstract \| Publisher Full Text\n\nEllegren H: Microsatellites: simple sequences with complex evolution. Nat Rev Genet. 2004; 5(6): 435–45. PubMed Abstract \| Publisher Full Text\n\nJi YJ, Zhang DX: Characteristics of microsatellite DNA in lepidopteran genomes and implications for their isolation. Acta Zool Sin. 2004; 50(4): 608–14. Reference Source\n\nMeglécz E, Petenian F, Danchin E, et al.: High similarity between flanking regions of different microsatellites detected within each of two species of Lepidoptera: Parnassius apollo and Euphydryas aurinia. Mol Ecol. 2004; 13(6): 1693–700. PubMed Abstract \| Publisher Full Text\n\nShah AB, Schielzeth H, Albersmeier A, et al.: High-throughput sequencing and graph-based cluster analysis facilitate microsatellite development from a highly complex genome. Ecol Evol. 2016; 6(16): 5718–27. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTay WT, Behere GT, Batterham P, et al.: Generation of microsatellite repeat families by RTE retrotransposons in lepidopteran genomes. BMC Evol Biol. 2010; 10(1): 144. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKalia RK, Rai MK, Kalia S, et al.: Microsatellite markers: an overview of the recent progress in plants. Euphytica. 2011; 177(3): 309–34. Publisher Full Text\n\nPannunzio NR, Watanabe G, Lieber MR: Nonhomologous DNA end-joining for repair of DNA double-strand breaks. J Biol Chem. 2018; 293(27): 10512–23. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWilder J, Hollocher H: Mobile elements and the genesis of microsatellites in dipterans. Mol Biol Evol. 2001; 18(3): 384–92. PubMed Abstract \| Publisher Full Text\n\nLiu R, Koyanagi KO, Chen S, et al.: Evolutionary force of AT-rich repeats to trap genomic and episomal DNAs into the rice genome: lessons from endogenous pararetrovirus. Plant J. 2012; 72(5): 817–28. PubMed Abstract \| Publisher Full Text\n\nKim KY, Kang PD, Lee KG, et al.: Microsatellite analysis of the silkworm strains (Bombyx mori): high variability and potential markers for strain identification. Genes Genomics. 2010; 32(6): 532–43. Publisher Full Text\n\nKim KY, Kang PD, Kim MJ, et al.: Microsatellite analysis of silkworm strains (Bombyx mori) of japan origin preserved in Korea. Int J Ind Entomol. 2014; 28(2): 39–50. Publisher Full Text\n\nHou CX, Li MW, Zhang YH, et al.: Analysis of SSR Fingerprints in Introduced Silkworm Germplasm Resources. Agric Sci China. 2007; 6(5): 620–7. Publisher Full Text\n\nChandrakanth N, Moorthy SM, Anusha P, et al.: Evaluation of genetic diversity in silkworm (Bombyx mori L.) strains using microsatellite markers. Int J Biotechnol Allied Fields. 2014; 2(3): 73–93.\n\nLi MW, Shen L, Xu AY, et al.: Genetic diversity among silkworm (Bombyx mori L., Lep., Bombycidae) germplasms revealed by microsatellites. Genome. 2005; 48(5): 802–10. PubMed Abstract \| Publisher Full Text\n\nKhurad AM, Zhang MJ, Deshmukh CG, et al.: A new continuous cell line from larval ovaries of silkworm, Bombyx mori. In Vitro Cell Dev Biol Anim. 2009; 45(8): 414–9. PubMed Abstract \| Publisher Full Text\n\nKhurad AM, Bahekar RS, Zhang MJ, et al.: Development and characterization of a new Bombyx mori cell line for protein expression. J Asia Pac Entomol. 2013; 16(1): 17–22. Publisher Full Text\n\nXu M, Tan J, Wang X, et al.: Establishment and characterization of a new embryonic cell line from the silkworm, Bombyx mori. ISJ. 2015; 12(1): 13–8. Reference Source\n\nMcIntosh AH, Grasela JJ, Matteri RL: Identification of insect cell lines by DNA amplification fingerprinting (DAF). Insect Mol Biol. 1996; 5(3): 187–95. PubMed Abstract \| Publisher Full Text\n\nSchugl MD, Mackay TF, Aquadro CF: Low mutation rates of microsatellite loci in Drosophila melanogaster. Nat Genet. 1997; 15(1): 99–102. PubMed Abstract \| Publisher Full Text\n\nBindroo BB, Moorthy SM: Genetic divergence, implication of diversity, and conservation of silkworm, Bombyx mori. Int J Biodivers. 2014; 2014: 15. Publisher Full Text\n\nSolignac M, Vautrin D, Baudry E, et al.: A microsatellite-based linkage map of the honeybee, Apis mellifera L. Genetics. 2004; 167(1): 253–62. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWatanabe TK, Ono T, Okuno S, et al.: Characterization of newly developed SSLP markers for the rat. Mamm Genome. 2000; 11(4): 300–5. PubMed Abstract \| Publisher Full Text\n\nXia QY, Guo YR, Zhang Z, et al.: Complete resequencing of 40 genomes reveals domestication events and genes in silkworm (Bombyx). Science. 2009; 326(5951): 433–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEstoup A, Jarne P, Cornuet JM: Homoplasy and mutation model at microsatellite loci and their consequences for population genetics analysis. Mol Ecol. 2002; 11(9): 1591–604. PubMed Abstract \| Publisher Full Text\n\nDe Barba M, Miquel C, Lobréaux S, et al.: High-throughput microsatellite genotyping in ecology: improved accuracy, efficiency, standardization and success with low-quantity and degraded DNA. Mol Ecol Resour. 2016; 17(3): 492–507. PubMed Abstract \| Publisher Full Text\n\nZhang L, Huang YP, Miao XX, et al.: Microsatellite markers application on domesticated silkworm and wild silkworm. Insect Sci. 2005; 12(6): 413–9. Publisher Full Text\n\nKômoto N, Tsuda M, Okada E, et al.: Development of methods for risk assessment of transgenic silkworms rearing on biodiversity. J Insect Biotechnol Sericology. 2014; 83(2): 171–9. [in Japanese, English abstract]. Publisher Full Text\n\nQian HY, Xu AY, Zhang YH, et al.: Genetic diversity and molecular phylogenetics on silkworm, Bombyx mori. Journal-Shenyang Agric Univ. 2007; 38(3): 357–61. [in Chinese, English abstract].\n\nThiyagu T, Kamble CK: DNA profiling of bivoltine silkworm germplasm races through microsatellite markers. Int J Biotechnol. 2011; 4(2): 53–7.\n\nVijayan K, Nair CV, Urs SR: Assessment of genetic diversity in the tropical mulberry silkworm (Bombyx mori L.) with mtDNA-SSCP and SSR markers. Emir J Food Agr. 2010; 22(2): 71–83. Publisher Full Text\n\nChu Q, Peng Y: Analysis of genetic relationship between color cocoon silkworm varieties by SSR markers. Hunan Agric Sci. 2013; 17: 13–6.\n\nRaju PJ, Krishnamurthy NB: Breeding of two bivoltines, MG511 and MG512, of silkworm, Bombyx mori L., for higher viability and silk productivity. Sericologia (France). 1993; 33(4): 577–93.\n\nDas SK, Sen SK, Saratchandra B: Improvement of commercial traits in mulberry silkworm Bombyx mori L by hybridization, Persp. Cytol Genet. 1998; 9: 101–8.\n\nDas SK: Techniques of breeding for evolving improved breeds of bivoltine mulberry silkworm Bombyx mori. Cytol Genet. 2001; 10: 129–34.\n\nMoorthy SM, Das SK, Kar NB, et al.: Breeding of Bivoltine breeds of Bombyx mori L suitable for variable climatic conditions of the tropics. Int J Ind Entomol. 2007; 14(2): 99–105. Reference Source\n\nLu C, Yu HS, Xiang ZH: Molecular systematic studies on Chinese mandarina silkworm (Bombyx mandarina M.) and domestic silkworm (Bombyx mori L.). Agric Sci China. 2002; 1(3): 349–58. [in Chinese, English abstract].\n\nDalirsefat SB, Mirhoseini SZ: Assessing genetic diversity in Iranian native silkworm (Bombyx mori L.) strains and Japanese commercial lines using AFLP markers. Iran J Biotechnol. 2007; 5(1): 25–33. Reference Source\n\nSorenson MD, DaCosta JM: Genotyping HapSTR loci: phase determination from direct sequencing of PCR products. Mol Ecol Resour. 2011; 11(6): 1068–75. PubMed Abstract \| Publisher Full Text\n\nNagaraja GM, Mahesh G, Satish V, et al.: Genetic mapping of Z chromosome and identification of W chromosome-specific markers in the silkworm, Bombyx mori. Heredity (Edinb). 2005; 95(2): 148–57. PubMed Abstract \| Publisher Full Text\n\nMiao XX, Li WH, Li MW, et al.: Inheritance and linkage analysis of co-dominant SSR markers on the Z chromosome of the silkworm (Bombyx mori L.). Genet Res (Camb). 2008; 90(2): 151–6. PubMed Abstract \| Publisher Full Text\n\nSperling F: Sex-linked genes and species differences in Lepidoptera. Can Entomol. 1994; 126(3): 807–18. Publisher Full Text\n\nHou CX, Sun PJ, Guo XJ, et al.: Marker-assisted selection in breeding silkworm strains with high silk production and resistance to the densonucleosis virus. Genet Mol Res. 2013; 12(4): 4171–8. PubMed Abstract \| Publisher Full Text\n\nLi MW, Hou CX, Zhao YP, et al.: Detection of homozygosity in near isogenic lines of non-susceptible to Zhenjiang strain of densonucleosis virus in silkworm. Afr J Biotechnol. 2007; 6(14): 1629–1633. Reference Source\n\nGao R, Li CL, Tong XL, et al.: Insight into genetic basis of Bombyx mori resistant strains with resistance to BmNPV by molecular linkage analysis. Sci Agric Sin. 2017; 50(1): 195–204. Publisher Full Text\n\nXu HM, Wei CS, Tang YT, et al.: A new mapping method for quantitative trait loci of silkworm. BMC Genet. 2011; 12(1): 19. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLi C, Zuo W, Tong X, et al.: A composite method for mapping quantitative trait loci without interference of female achiasmatic and gender effects in silkworm, Bombyx mori. Anim Genet. 2015; 46(4): 426–32. PubMed Abstract \| Publisher Full Text\n\nLi M, Guo Q, Hou C, et al.: Linkage and mapping analyses of the densonucleosis non-susceptible gene nsd-Z in the silkworm Bombyx mori using SSR markers. Genome. 2006; 49(4): 397–402. PubMed Abstract \| Publisher Full Text\n\nMiao XX, Li MW, Dai FY, et al.: Linkage analysis of the visible mutations Sel and Xan of Bombyx mori (Lepidoptera: Bombycidae) using SSR markers. Eur J Entomol. 2007; 104(4): 647–52. Publisher Full Text\n\nXiang H, Li M, Yang F, et al.: Fine mapping of Ekp-1, a locus associated with silkworm (Bombyx mori) proleg development. Heredity (Edinb). 2008; 100(5): 533–40. PubMed Abstract \| Publisher Full Text\n\nZhao YP, Li MW, Xu AY, et al.: SSR based linkage and mapping analysis of C, a yellow cocoon gene in the silkworm, Bombyx mori. Insect Sci. 2008; 15(5): 399–404. Publisher Full Text\n\nLi X, Li MW, Guo QH, et al.: Mapping of the yellow inhibitor gene I in silkworm Bombyx mori using SSR markers. Yi Chuan. 2008; 30(8): 1039–42. PubMed Abstract \| Publisher Full Text\n\nBai HC, Xu AY, Li MW, et al.: SSR based linkage and mapping analysis of dominant endurance to fluoride gene (Def) in the silkworm, Bombyx mori. Acta Seric Sin. 2008; 34(2): 191–6. Reference Source\n\nDai FY, Qiao L, Tong XL, et al.: Mutations of an arylalkylamine-N-acetyltransferase, Bm-iAANAT, are responsible for silkworm melanism mutant. J Biol Chem. 2010; 285(25): 19553–60. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhan S, Guo QH, Li MH, et al.: Disruption of an N-acetyltransferase gene in the silkworm reveals a novel role in pigmentation. Development. 2010; 137(23): 4083–90. PubMed Abstract \| Publisher Full Text\n\nWang XY, Li MW, Zhao YP, et al.: Mapping of non-lepis wing gene nlw in silkworm (Bombyx mori) using SSR and STS markers. Yi Chuan. 2010; 32(1): 54–8. PubMed Abstract \| Publisher Full Text\n\nZhao Y, Zhang J, Wu YC, et al.: SSR marker-based mapping and linkage analysis of Bombyx mori thermo tolerance gene. J Food Agric Environ. 2010; 8(1): 338–42. Reference Source\n\nZhao XM, Wei GQ, Liu CL, et al.: Linkage and mapping analyses of the no glue egg gene Ng in the silkworm (Bombyx mori L.) using simple sequence repeats (SSR) markers. African J Biotechnol. 2011; 10(47): 9549–56. Publisher Full Text\n\nChen P, Tong XL, Li DD, et al.: Antennapedia is involved in the development of thoracic legs and segmentation in the silkworm, Bombyx mori. Heredity (Edinb). 2013; 111(3): 182–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChen P, Tong XL, Li DD, et al.: Fine mapping of a supernumerary proleg mutant (ECs-l) and comparative expression analysis of the abdominal-A gene in silkworm, Bombyx mori. Insect Mol Biol. 2013; 22(5): 497–504. Publisher Full Text\n\nChen A, Gao P, Zhao QL, et al.: Mutation of a vitelline membrane protein, BmEP80, is responsible for the silkworm “Ming” lethal egg mutant. Gene. 2013; 515(2): 313–9. PubMed Abstract \| Publisher Full Text\n\nWei GQ, Yu L, Liu CL, et al.: Linkage and mapping analyses of the normal marking gene +P in the silkworm (Bombyx mori) using SSR markers. Genet Mol Res. 2013; 12(3): 2351–9. PubMed Abstract \| Publisher Full Text\n\nTong XL, He S, Chen J, et al.: A novel laminin β gene BmLanB1-w regulates wing-specific cell adhesion in silkworm, Bombyx mori. Sci Rep. 2015; 5: 14. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang RX, Tong XL, Gai TT, et al.: A serine protease homologue Bombyx mori scarface induces a short and fat body shape in silkworm. Insect Mol Biol. 2018; 27(3): 319–32. PubMed Abstract \| Publisher Full Text" }	[ { "id": "54605", "date": "28 Oct 2019", "name": "Nalavadi Chandrakanth", "expertise": [ "Reviewer Expertise Silkworm Breeding and Biotechnology" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors have scripted the manuscript properly and covered the appropriate and relevant literature necessary for understanding the characteristics of microsatellite and the evolutionary events associated with it in Bombyx mori genome. Further this review has focused on the literature pertaining to the applications of the microsatellite as markers, in genetic analysis studies and in construction of the linkage maps in B. mori. Interestingly, the authors have also discussed about the lacuna of using microsatellite in marker-trait association studies and construction of high density linkage maps in B. mori. The authors have brought out an important point that linkage map construction in silkworm cannot be adjusted to the other models due to their gender based effects and absence of chromosomal cross-linking in the germinal line of the female silkworms. In this line, recently proposed statistical models to analyze QTLs were also reported and explained in this manuscript.\n\nComments\n\nIn the last paragraph of Introduction, it was mentioned that this review will focus on the roles or strategies involving microsatellite markers for conservation of silkworm germplasm. But, in future scope, this part has to be included with effective strategies to conserve the silkworm germplasm using microsatellite markers. In addition, the authors can also report in the future scope about how these microsatellite markers has been employed in breeding programs to improve the genetic materials of different silkworms.\n\nIn some parts of the manuscript, the sentences are very lengthy which can be simplified.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Yes", "responses": [] }, { "id": "59560", "date": "24 Feb 2020", "name": "Maria-Lucia Carneiro Vieira", "expertise": [ "Reviewer Expertise Plant genetics and genomics" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI do believe that the review should be useful for readers, it considers the state of the art on the role of microsatellitesfor the genetic analysis of Bombyx mori silkworm. The most recent citation dates from 2018; therefore, I suggest authors to search for more recent articles on other Lepidoptera of agronomic importance. The population of these pests tend to have the same genetic structure as the ones of Bombyx mori silkworm? Please make the comparison.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1424
https://f1000research.com/articles/7-743/v1	14 Jun 18	{ "type": "Software Tool Article", "title": "CyTargetLinker app update: A flexible solution for network extension in Cytoscape", "authors": [ "Martina Kutmon", "Friederike Ehrhart", "Egon L. Willighagen", "Chris T. Evelo", "Susan L. Coort", "Friederike Ehrhart", "Egon L. Willighagen", "Chris T. Evelo", "Susan L. Coort" ], "abstract": "Here, we present an update of the open-source CyTargetLinker app for Cytoscape (http://apps.cytoscape.org/apps/cytargetlinker) that introduces new automation features. CyTargetLinker provides a simple interface to extend networks with links to relevant data and/or knowledge extracted from so-called linksets. The linksets are provided on the CyTargetLinker website or can be custom-made for specific use cases. The new automation feature enables users to programmatically execute the app’s functionality in Cytoscape (command line tool) and with external tools (e.g. R, Jupyter, Python, etc). This allows users to share their analysis workflows and therefore increase repeatability and reproducibility. Three use cases demonstrate automated workflows, combinations with other Cytoscape apps and core Cytoscape functionality. We first extend a protein-protein interaction network created with the stringApp, with compound-target interactions and disease-gene annotations. In the second use case, we created a workflow to load differentially expressed genes from an experimental dataset and extend it with gene-pathway associations. Lastly, we chose an example outside the biological domain and used CyTargetLinker to create an author-article-journal network for the five authors of this manuscript using a two-step extension mechanism. With 300 downloads per month in the last year and over 12,000 downloads in total, CyTargetLinker shows the adoption and relevance of the app in the field of network biology. In April 2018, the original publication was cited in 57 articles demonstrating the applicability in biomedical research.", "keywords": [ "Cytoscape", "CyTargetLinker", "network extension", "network visualization", "regulatory networks", "data integration" ], "content": "Introduction\n\nThe Cytargetlinker app provides a flexible and simple way to extend networks in Cytoscape1 with links to (prior) knowledge from external sources. Since its first release in 20132, CyTargetLinker has been downloaded more than 12,000 times and used in numerous studies. These applications in biological studies range from the creation of a microRNA-gene association network for lipid diseases3 or Alzheimer’s disease4 to the application of algorithms for drug sensitivity prediction5.\n\nWhile the app was originally intended to be used for the extension of biological networks with regulatory interactions, researchers have used CyTargetLinker to integrate knowledge about many different types of relationships (e.g. pathway associations and disease annotations). Therefore, we renamed the previously used Regulatory Interaction Networks (RegINs) to linksets to make the broader applicability more explicit. Moreover, the generation of linksets, either manually or in an automated manner, has become more user-friendly.\n\nIn this new version of CyTargetLinker, we introduce an automation feature that allows programmatic access to the app functionality. In the Results section, we present three use cases that highlight the app’s purpose, how it can be easily combined with other Cytoscape apps and the advantages of the automation. Whereas the first two use cases have a biological nature, the third use case demonstrates the broader applicability with a non-biological example.\n\n\nMethods\n\nThe newest version of CyTargetLinker (4.0.0+) was developed for Cytoscape (3.6.0+) which introduces a new interface for automation that can make apps callable as services by the Cytoscape Command scripts, Python and R. This promotes open and reproducible data analysis, and simple integration with other apps. CyTargetLinker can be installed through the Cytoscape app store.\n\nOn the CyTargetLinker website, we provide a variety of linksets for regulatory interactions, pathway associations and disease annotations (https://projects.bigcat.unimaas.nl/cytargetlinker/linksets/). Additionally, we deliver a simple Java program to convert tab delimited text files into XGMML linksets that can be used with CyTargetLinker (https://github.com/CyTargetLinker/linksetCreator). Using BridgeDb6, a framework for finding and mapping database identifiers, the script enables the support of multiple identifier systems for biological entities.\n\nWhile CyTargetLinker can still be used through the Cytoscape graphical user interface (see online tutorials), we would like to highlight the novel application programming interface (API) that allows the programmatic execution of the app’s functionality.\n\nCyTargetLinker provides a set of API methods to automise the extension of networks, see Table 1. The key function is the “extend” function, which parses the provided LinkSets and extracts relevant interactions for the selected network. The user can then choose to use the CyTargetLinker visual style and the force-directed layout. Often, users want to integrate knowledge for the same interaction type from different resources. With the “filterOverlap” function, users can visualise only those interactions that are supported by multiple resources.\n\nList of the API methods of CyTargetLinker, their parameters and a general description.\n\n\nUse cases\n\nThe broad applicability of CyTargetLinker will be demonstrated in three different use cases. The focus lies on the automation of the analysis and the R scripts for each use case are provided in the tutorial repository on GitHub. We chose to present two biological and one non-biological use cases to demonstrate the flexibility of the app.\n\nUse case 1: Investigating drug-targets and disease associations for a Rett syndrome protein-protein interaction network. Rett syndrome is a rare disease caused by a mutation in the methyl-CpG-binding protein 2 (MECP2) gene7. In this use case, we used the stringApp8 of Cytoscape to create a protein-protein interaction (PPI) network for Rett syndrome (Disease Query). The PPI network is then extended using CyTargetLinker with compound-target interactions from ChEMBL9,10 and disease-gene associations from a manually curated subset for rare diseases from OMIM11. ChEMBL is an open online bioactivity database containing information about compounds, their bioactivity and their possible targets (including proteins). OMIM is a comprehensive collection of human genetic phenotypes and their associated human genes. Both linksets are available on the linkset download page on the CyTargetLinker website.\n\nFirst, the stringApp was used to create a Rett syndrome PPI (query=“Rett syndrome”, cutoff=0.4, limit=20). Using CyTargetLinker, the network was extended with 37 compound-target interactions from ChEMBL and 18 gene-disease associations from OMIM (see Figure 1).\n\nThe protein-protein interaction network for Rett syndrome was created using the disease query option of the stringApp for Cytoscape. The proteins are represented as gray circles. Then, CyTargetLinker was used to extend the network with compounds from ChEMBL (purple diamonds) and disease annotations from OMIM (blue octagons).\n\nUse case 2: Pathway associations for differentially expressed genes in Rett syndrome. For this use case, we selected a list of differentially expressed genes in the Purkinje cells located in the cerebellar cortex of the brain of a Mecp2−/y mouse model12,13 for Rett syndrome. Next, we investigated in which biological processes these altered genes are involved. Using the pathway annotations from the WikiPathways database14, CyTargetLinker adds the pathway information and creates a pathway-gene network.\n\nFrom the dataset, we extracted 65 genes with an absolute log2 fold change larger than 1. Only 16 genes are present in one or more pathways of the curated mouse pathway collection from WikiPathways. Figure 2 shows the resulting gene-pathway network. Genes without pathway annotations have been removed. Differential gene expression is shown on the gene nodes (blue = down, red = up) and green border color of the pathway nodes indicates that the pathway has been identified as significantly affected through overrepresentation analysis in the pathway analysis tool PathVisio15.\n\nDEGs in a mouse model for Rett syndrome (Mecp2−/y mice) were selected and imported in Cytoscape (circular nodes). The genes are colored based on changes in gene expression (blue=down-regulated, red=up-regulated). Thereafter, gene-pathway associations from WikiPathways were added and the pathways are shown as gray rectangles. Genes without pathway annotations have been removed. The green border color indicates pathways that are significantly altered based on over-representation analysis in PathVisio (Z-Score > 1.96).\n\nUse case 3: Author-publication-journal network. This example uses two custom made linksets for author-article and article-journal relationships from Wikidata16–18. After loading the initial five author nodes in Cytoscape, we performed a two-step extension with CyTargetLinker. We first added publications from the author-article linkset and then the journals from the article-journal linkset, see Figure 3. Author nodes are colored in gray, articles in yellow and journals in green. The network clearly shows the collaborations and diversity between the authors. Layout and visual style was slightly adapted manually in the graphical user interface to improve the readability of the network.\n\nFor the five authors of this manuscript, two custom linksets were used to first add author-article relationships and in a second step add article-journal relationships.\n\n\nDiscussion\n\nOne of the major challenges in science is the reproducibility of results presented in articles. Besides the challenges in reproducibility of experiments, the computational analyses are also often unclear and insufficiently described19. Automation of analysis workflows enables researchers to share the details of their computational analyses and enables simple reproducibility of the results.\n\nHere, we introduce the new version of the CyTargetLinker app, which provides full programmatic execution of the functions from within Cytoscape (command line tool) and external tools (e.g. R, Jupyter, Python, etc). The network extension is therefore reproducible and repeatable with other input data. Consequently, users can build scripts that run common analysis workflows and combine CyTargetLinker with other apps, as shown in Use case 1 (stringApp). The integration of CyTargetLinker in Cytoscape gives access to a powerful set of visualization options, as demonstrated in Use case 2.\n\nAs part of the Cytoscape tutorials collection for online presentations, we developed a CyTargetLinker tutorial presentation using Reveal.js. This tutorial presentation can be reused and adapted for specific teaching activities. Together with our tutorials for the Cytoscape desktop application and the automation example scripts, relevant documentation for users is provided to get familiar with the functionality of CyTargetLinker.\n\nThe generic nature of CyTargetLinker has been highlighted by renaming RegINs to linksets, and we now provide a variety of different linksets on our website. The XGMML structure of the linksets is simple, instructions how to create them from tab-delimited text files are available, and CyTargetLinker could therefore be used for non-biological networks as well (shown in use case 3).\n\n\nConclusions\n\nIn this paper, we showcase the latest update of the CyTargetLinker app for Cytoscape and its new automation feature. The ability to programmatically execute the app’s functionality opens up the possibility to build complex workflows that are repeatable and reproducible. We also explored the broader applicability of the app besides the originally intended use for regulatory network extension. Due to the flexible design of the app and the linksets, we are now also showcasing other use cases, including non-biological networks.\n\n\nData availability\n\nLinksets, tutorials and link to source code (app and linkset creator) are available from the CyTargetlinker app website: https://projects.bigcat.unimaas.nl/cytargetlinker/.\n\n\nSoftware availability\n\n1. The app is available from the Cytoscape app store: http://apps.cytoscape.org/apps/cytargetlinker.\n\n2. Link to source code: https://github.com/CyTargetLinker/cytargetlinker.\n\n3. Archived source code at time of publication: https://doi.org/10.5281/zenodo.121872120.\n\n4. Tutorials including R code for use cases: https://github.com/CyTargetLinker/cytargetlinker-tutorials.\n\n5. Software license: https://www.apache.org/licenses/LICENSE-2.0.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis project has been co-financed by the Dutch Province of Limburg and ELIXIR, the European research infrastructure for life-science data.\n\n\nAcknowledgements\n\nWe would like to thank the Cytoscape developer team, especially Barry Demchak, Scooter Morris and Alex Pico, for the support during the automation feature implementation.\n\n\nReferences\n\nShannon P, Markiel A, Ozier O, et al.: Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003; 13(11): 2498–2504. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKutmon M, Kelder T, Mandaviya P, et al.: CyTargetLinker: a cytoscape app to integrate regulatory interactions in network analysis. PLoS One. 2013; 8(12): e82160. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKandhro AH, Shoombuatong W, Nantasenamat C, et al.: The MicroRNA Interaction Network of Lipid Diseases. Front Genet. 2017; 8: 116. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMoradifard S, Hoseinbeyki M, Ganji SM, et al.: Analysis of microRNA and Gene Expression Profiles in Alzheimer's Disease: A Meta-Analysis Approach. Sci Rep. 2018; 8(1): 4767. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDe Niz C, Rahman R, Zhao X, et al.: Algorithms for drug sensitivity prediction. Algorithms. 2016; 9(4): 77. Publisher Full Text\n\nvan Iersel MP, Pico AR, Kelder T, et al.: The BridgeDb framework: standardized access to gene, protein and metabolite identifier mapping services. BMC Bioinformatics. 2010; 11(1): 5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeonard H, Cobb S, Downs J: Clinical and biological progress over 50 years in Rett syndrome. Nat Rev Neurol. 2017; 13(1): 37–51. PubMed Abstract \| Publisher Full Text\n\nSzklarczyk D, Morris JH, Cook H, et al.: The STRING database in 2017: quality-controlled protein-protein association networks, made broadly accessible. Nucleic Acids Res. 2017; 45(D1): D362–D368. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBento AP, Gaulton A, Hersey A, et al.: The ChEMBL bioactivity database: an update. Nucleic Acids Res. 2014; 42(Database issue): D1083–D1090. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWillighagen EL, Waagmeester A, Spjuth O, et al.: The ChEMBL database as linked open data. J Cheminform. 2013; 5(1): 23. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHamosh A, Scott AF, Amberger JS, et al.: Online Mendelian Inheritance in Man (OMIM), a knowledgebase of human genes and genetic disorders. Nucleic Acids Res. 2005; 33(Database issue): D514–D517. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSugino K, Hempel CM, Okaty BW, et al.: Cell-type-specific repression by methyl-CpG-binding protein 2 is biased toward long genes. J Neurosci. 2014; 34(38): 12877–12883. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEhrhart F, Coort SL, Cirillo E, et al.: New insights in rett syndrome using pathway analysis for transcriptomics data. Wien Med Wochenschr. 2016; 166(11–12): 346–352. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSlenter DN, Kutmon M, Hanspers K, et al.: WikiPathways: a multifaceted pathway database bridging metabolomics to other omics research. Nucleic Acids Res. 2017; 46(D1): D661–D667. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKutmon M, van Iersel MP, Bohler A, et al.: PathVisio 3: an extendable pathway analysis toolbox. PLoS Comput Biol. 2015; 11(2): e1004085. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVrandečić D, Krötzsch M: Wikidata: a free collaborative knowledgebase. Communications of the ACM. 2014; 57(10): 78–85. Publisher Full Text\n\nNielsen FÅ, Mietchen D, Willighagen E: Scholia, scientometrics and wikidata. In Lecture Notes in Computer Science. Springer International Publishing, 2017; 237–259. Publisher Full Text\n\nTaraborelli D, Mietchen D: Wikicite 2017 report. 2017. Publisher Full Text\n\nReality check on reproducibility. Nature. 2016; 533(7604): 437. PubMed Abstract \| Publisher Full Text\n\nKutmon M: CyTargetLinker: Release 4.0.1. 2018. Data Source" }	[ { "id": "35060", "date": "28 Jun 2018", "name": "Alex Gutteridge", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe CyTargetLinker Cytoscape app presented here provides a useful mechanism for extending existing biological networks. The rationale for this functionality is well explained by the authors and the general utility is demonstrated by the citation of several previous studies using the tool and the variety of use cases presented.\nThe API interface, which is the focus of this update, is a very useful addition which allows the scripting and hence reproducibility of this kind of analysis. The API is simple, but covers the fundamental use case of the app. I can only find partial documentation of the API method parameters however. They are named in Table 1, but I could not find more extensive documentation on the app website (https://projects.bigcat.unimaas.nl/cytargetlinker/) which mostly refers to an earlier version of the app from what I can tell. The API methods should be documented somewhere more completely. This is why I have answered 'Partly' to 'Are sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others?'.\nThe interpretation of the results and conclusions about the tool are straight forward. I have not tested the performance of the app outside the use cases the authors provide, but in those contexts performance is perfectly adequate. Perhaps the authors could comment on whether the app has performance limits on the size of network that can be annotated in this way and how that compares to Cytoscape's underlying limits.\nI tested the first use case script with Cytoscape 3.6.1, stringApp 1.1.1 and CyTargetLinker 4.0.1 (the latest versions at time of writing). This revealed a small API incompatibility. The script attempts to link on idAttribute 'display name', but no such attribute comes from this version of stringApp. Changing the idAttribute to 'name' allows the script to work. I can't find which version of stringApp the authors are using, but maybe it needs to be specified or the script updated. This is why I have marked 'Partly' to the question 'Are sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others?' Otherwise the use case reproduced fine.\nI have not tested the CyTargetLinker Java program, but it seems reasonably well documented and straight forward to run.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4837", "date": "13 Aug 2019", "name": "Martina Summer-Kutmon", "role": "Author Response", "response": "Thank you for your comments. We completely restructured and updated the website and all tutorials and example scripts should now contain the relevant information to reproduce the results." } ] }, { "id": "40334", "date": "05 Dec 2018", "name": "John H. Morris", "expertise": [ "Reviewer Expertise Network Biology" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors present an update to their successful CyTargetLinker app for Cytoscape. This update extends the basic capabilities of CyTargetLinker by generalizing some concepts and adding automation functionality. The article is clear and well-written and succinctly outlines the major point of the article and the features the article is going to focus on. They present three use cases, which outline the automation capabilities nicely.\nI did have several issues that I believe should be addressed to support users trying to follow the instructions and to elucidate more clearly the functionality of the new automation features.\nFirst, the instructions point to a GitHub repository for the three use cases, however, I could only find two use cases in the repository. This should absolutely be fixed before approval.\nSecond, the authors state that both the ChEMBL and OMIM linksets are available on the CyTargetLinker website. Unfortunately, only the ChEMBL linkset is provided. The OMIM linkset is provided in the github repository, however, once the file is unzipped. It wasn't clear to me why the authors chose to provide the data as a zip file in UseCase1, but just provide the unzipped directories in UseCase2. I would suggest just to leave the files unzipped in UseCase1 in the same way you do for UseCase2.\nThird, to support general understanding of the automation, I would add the CyTargetLinker commands you actually execute to achieve the results. For UseCase1, this would be:\n\"cytargetlinker extend idAttribute=\\\"display name\\\" linkSetFiles=\\\"/tmp/cytargetlinker-tutorials/R/UseCase1/LinkSets/chembl_23_hsa_20180126.xgmml,/tmp/cytargetlinker-tutorials/R/UseCase1/LinkSets/rare-disease-gene-associations-hsa-20180411.xgmml\\\" network=current\". Obviously, the paths could be truncated, so something like: \"cytargetlinker extend idAttribute=\\\"display name\\\" linkSetFiles=\\\"path/chembl_23_hsa_20180126.xgmml,path/rare-disease-gene-associations-hsa-20180411.xgmml\\\" network=current\" This would make it much easier for app writers to utilize your command features without having to look through the R source code.\nAnd, as mentioned before, I couldn't find UseCase3.\nFinally, as a suggestion, it would be nice to have the functionality to create a linkset from a csv file as part of the app so the user wouldn't need to download and execute it separately.\nOverall, I'm enthusiastic about the app and the new capabilities, and would be happy to approve it after the above major changes are made.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4838", "date": "13 Aug 2019", "name": "Martina Summer-Kutmon", "role": "Author Response", "response": "Thank you for your comments. We completely restructured and updated the website. We also made sure the third use case script is available in the automation repository. We provided the data for use case 1 as a zip file because the ChEMBL linkset is too large to be uploaded to Github. I now only zipped that specific linkset and gave instructions in the R script to first unzip that specific xgmml file. This makes it more consistent with the other use cases. We added the general commands used to execute the extension for each of the use cases to the manuscript. Since we are using the linkset creator script to generate all linksets regularly, we opted for keeping it separate. Most users use the linksets provided or contact us to add a new one and only few generate their own. We hope that we addressed all of your requests in version 2 and the new updated website." } ] } ]	1	https://f1000research.com/articles/7-743
https://f1000research.com/articles/8-711/v1	22 May 19	{ "type": "Method Article", "title": "Identification of nitrite treated tuna fish meat via the determination of nitrous oxide by head space-gas chromatography/mass spectrometry", "authors": [ "Markus Niederer", "Sandra Lang", "Bernard Roux", "Thomas Stebler", "Christopher Hohl", "Sandra Lang", "Bernard Roux", "Thomas Stebler", "Christopher Hohl" ], "abstract": "Tuna fish meat is an expensive and highly perishable sea food. Fresh meat has a bright red colour which soon turns into an unsightly brown during storage. To prolong the aspect of freshness, the red colour is stabilised or even enhanced e.g. with carbon monoxide or nitric oxide, the product of a nitrite / ascorbic acid treatment, which bind as a ligand to myoglobin. These procedures are illegal. Here we present a method for identifying tuna meat samples, which have undergone fraudulent wet salting with nitrite. The method uses headspace-gas chromatography/mass spectrometry (GC/MS) for the determination of nitrous oxide, which is formed as the final product of the two-step reduction nitrite (added agent) to nitric oxide (ligand) to nitrous oxide (target compound). Complex bound nitric oxide is set free with sulfuric acid, which also promotes the reduction to nitrous oxide. The method was validated using 15N labelled nitrite as well as treated and untreated reference fish samples. A survey of 13 samples taken from the Swiss market in 2019 showed that 45 % of all samples were illegally treated with nitrite.", "keywords": [ "tuna", "red colour stability", "food fraud", "nitrite", "nitrous oxide", "food additives" ], "content": "Introduction\n\nAs Tuna fish species have been overfished on the one hand, and their meat being a sought-after but highly perishable food stuff on the other hand, the seafood industry has felt itself constrained to undertake efforts in increasing the shelf life of this commodity. Being a bright red in its fresh state, but turning into an unsightly brown during storage, attempts have also been made in extending or even enhancing the attractive aspect of fresh meat. The red colour of meat in its natural state is due to myoglobin in its reduced form (Mb-Fe+II-O2). Being susceptible to autoxidation, the resulting metmyoglobin (Mb-Fe+III-O2) is brown in colour. One way of stabilising the red colour is by gassing with carbon monoxide (CO), which binds as a ligand to myoglobin (Mb-Fe+II-CO). A second way consists in applying the anion nitrite (NO2-) in combination with the reducing agent ascorbic acid to form nitric oxide (NO), which forms a complex with both myoglobin (Mb- Fe+II-NO) and metmyoglobin (Mb-Fe+III-NO)1. Whereas the colour hue of the CO complex is described as cherry red, that of the NO-complexes are described as pink red. Neither procedure is authorised for fresh fish2. However, with these possibilities at hand, abuse for simulating the fresh aspect of tuna fish meat is a tempting option3,4, all the more so regarding the currently insufficient tools for fraud detection: Whereas, for CO, analytical methods are at hand to support law enforcement5,6, attempts to detect nitrite treated meat, either by searching for nitrite residues or traces of nitrosamines, have not been reported as being successful yet7. Here we report on our method for the detection of nitrite treatment (brining), which originates from the method for CO determination and determines nitrous oxide (N2O) as the reduction product of nitric oxide (NO), stemming from the complex Mb- Fe+II-NO (Figure 1).\n\nIn bold: verification experiment with labelled Na15NO2 and resulting 15N2O.\n\n\nMethods\n\nSulphuric acid (95–97%, cat. no. 1.00731), 1-octanol (puriss., cat. no. 1.00991), sodium chloride (NaCl, p.a., cat. no.1.06404.1000), sodium nitrite (NaNO2, p.a., cat. no. 1.06549.0100), sodium nitrite-N15 (Na15NO2, 99%, cat. no. 490814-1G), ascorbic acid (p.a., cat. no. 127) were all from Merck (Buchs, Switzerland) and sodium ascorbate (99%, cat. no. 352681000) from Acros (New Jersey, U.S.A.). Certified NO- gas (100 ppmv N2, 5 L, 150 bar, cat. no.1878) was purchased from Carbagas (Basel, Switzerland) and N2O- gas (100 ppmv N2, 2 L, 150 bar, cat. no. 7092325) was from Messer AG (Lenzburg, Switzerland).\n\nTuna samples (frozen loins, tuna saku, loose and pre-prepared packages) were taken from retailers and importers in Basel and stored at – 18°C until sample preparation. An aliquot of 15 g was weighed into a 70 mL centrifuge tube and 15 g of crushed ice as well as 15 mL of water were added. The mixture was homogenised at 7,000–9,000 rpm with a polytron homogeniser (Kinematica AG, Switzerland) and centrifuged for 5 min at 3,500 rpm. An aliquot of the resulting supernatant (5 mL) was transferred into a 20 mL headspace vial. After adding 5 µL of n-octanol and 2 mL of sulphuric acid (20 % in deionised water) the vial was immediately sealed and gently shaken by hand for 30 s. Certified NO- gas 100 ppmv (122.6 ng/mL) and N2O- gas 100 ppmv in N2 (180.6 ng/mL) served as standards for calibration: A headspace vial was filled with 5 mL of deionised water and 2 mL of sulphuric acid (20 %), the cap only loosely fitted and the reference gas introduced with a needle through the septum. The cap was then sealed tight.\n\nThe sealed vials were placed into the incubation unit of the CombiPAL autosampler (CTC Analytics, Zwingen, Switzerland) for 60 min at 25°C and analysed with a Trace 1310 GC / ISQ quadrupole system (Thermo Scientific, U.S.A.) under the following conditions: 500 μL of headspace was injected at 70°C (split 1:20) onto a Porous Layer Open Tubular (PLOT) column (30 m × 0.32 mm inner diameter, 12 µm film thickness, HP-PLOT Molsieve, Agilent Technologies, cat. no. 19091P-MS4), Palo Alto, CA, U.S.A.). Helium (99.999%, Carbagas, Basel, Switzerland, cat. no. P6201RG) was used as carrier gas at a constant flow rate of 1 mL/min. The oven temperature was programmed from 70°C (5 min) to 220°C (3 min) at 30°C/min. The quadrupole detector was set in the electronic ionisation mode (full scan 10–100 amu) with selected-ion monitoring (70 eV, ion source = 170°C, delay = 2.5 min) with a transfer-line to the GC (1 m fused silica deactivated, 0.25 mm i.d.) at 290°C. Specific masses for the detection were m/z 30 for NO (m/z 31 for 15NO) at a retention time of about 7.9 minutes and m/z 44 for N2O (m/z 46 for 15N2O) at 11.4 minutes.\n\nFor brining experiments, 15 g of nine untreated fish samples (certified as 100 % tuna from the Maldives, the Philippines, China, Vietnam) were each cut into pieces of about 1 cm and treated with 20 mL of brine solution (10% NaCl and 0.4% NaNO2 in deionised water)8 at room temperature for 24 hours. After removing the solution, the now reddish looking tuna samples along with eight untreated control samples were each bathed twice in 15 mL of antioxidant solution (3.0 % ascorbic acid and sodium ascorbate in deionised water) for one hour at room temperature and subsequently washed twice with antioxidant solution. Thereafter, samples were prepared as described above and analysed by headspace-GC/MS.\n\nVerification experiment: The whole procedure was also done with three tuna samples using Na15NO2 for the brine solution instead.\n\n\nResults and discussion\n\nIn contrast to carbon monoxide, where the agent used for treatment, the ligand responsible for colour stabilisation and the target compound for sample screening is one and the same molecule, nitrite treatment identification has to consider the two-step reduction from nitrite to nitric oxide to nitrous oxide. This required changing the original GC temperature programme of the CO method: Instead of an isothermic run at 40°C, oven temperature was ramped up from 70°C to 220°C.\n\nBoth carbon monoxide and nitric oxide are bound in the sample as a ligand. Their release is enhanced by adding sulfuric acid. Being a free radical, however, nitric oxide is far more reactive than carbon monoxide: Instead of a sharp symmetrical peak, chromatograms show nitric oxide as a broad hump, barely distinguishable from baseline noise. The direct approach therefore is not feasible. In an acidic environment, however, two NO molecules, upon reduction, react to nitrous oxide (N2O), which shows up as a perfect peak in the chromatogram (Figure 2), allowing for the indirect detection of NO. For the purpose of identifying nitrite treatment, a semi quantitative assessment of N2O proved to be sufficient, as any clear signal is alone due to a former nitrite treatment.\n\nVerification of this mechanism was performed using isotope labelled nitrite (Na15NO2) for brining (Figure 1). During subsequent analysis, both the 15NO hump and 15N2O peak were detected (Figure 2), which is consistent with the postulated reaction pathway. In addition, brining experiments with nine tuna samples using unlabelled NaNO2 showed sharp N2O signals in the range of 1000 to 12000 μg/kg (Figure 3, underlying data9). In contrast, untreated reference samples, even though they were exposed to antioxidants, were free of nitrous oxide (limit of detection about 20 μg/kg). Therefore, naturally occurring relevant concentrations of N2O can be excluded.\n\n15N2O labelled samples (x).\n\nThe method was validated with seven reference tuna samples of a trustworthy source from 2017 which were stored at -18°C for two years in our lab (see underlying data10). Five of them were declared as untreated and consists of 100% tuna. The remaining two samples were declared as treated with rosemary extract (E392) and/or nitrite and labelled with additives such as salt and antioxidants (E300, E301, E331). The untreated samples were free of N2O whereas the treated ones showed stable N2O signals of 260 and 750 μg/kg. These results not only confirm our findings from the brining experiments, but are also consistent with the corresponding documents of the reference tuna samples.\n\nThe method was then used on a routine basis in a market survey of 13 samples of raw tuna collected in Basel in 2019 (see underlying data11). Seven samples (54%) were negative. Six samples (46%) were tested positive with N2O peaks corresponding to levels of between 90 and 1130 μg/kg. As a consequence, these six samples, all originating from Vietnam, were objected according to law.\n\n\nConclusion\n\nWith our experiments using labelled Na15NO2 and the comparison of our results with the corresponding documents of the reference tuna samples, we could demonstrate that our method using nitrous oxide as a target compound is suitable for identifying nitrite manipulated tuna. Its use in a market survey also showed, that the method is fit for purpose. In addition, the method was even capable of identifying nitrite treatment in samples stored at -18°C for two years before analysis.\n\n\nData availability\n\nAs a law enforcement body, we are bound to official secrecy. Considering this restriction, names of products and importers were anonymised.\n\nFigshare: Fig03 brining experiment rawdata.csv. https://doi.org/10.6084/m9.figshare.8142992.v19\n\nThis project contains the following underlying data:\n\nFig03 brining experiment rawdata.csv (Raw data and descriptive statistic data (univariate diagram) of \"figure 03\" performed with the Add-In XLSTAT 2009.1.02 is provided as Excel-file (CSV). The data include file name, sample name, type of treatment, area, calculated N2O amounts and statistical values)\n\nFigshare: Validation with Reference samples 2017. https://doi.org/10.6084/m9.figshare.8143016.v110\n\nThis project contains the following underlying data:\n\nValidation with Reference samples 2017.csv (Raw data of 7 reference samples of 2017 of the \"validation section\" are provided as Excel-file (CSV). The data include file name, sample name, certified type of treatment and calculated N2O amounts (single and average values))\n\nFigshare: Market survey 2019 rawdata. https://doi.org/10.6084/m9.figshare.8143031.v111\n\nThis project contains the following underlying data:\n\nMarket survey 2019 rawdata.csv (Raw data and descriptive statistic data of the \"market survey section\" performed with the Add-In XLSTAT 2009.1.02 are provided as Excel-file (CSV). The data include file name, sample name, area, calculated N2O amounts, test result and statistical values)\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nFaustman C: Myoglobin Chemistry and Modifications that Influence (Color and) Color Stability. In: American Meat Science Association, 67th Annual Reciprocal Meat Conference, 2014. Reference Source\n\nVerordnung des EDI über die zulässigen Zusatzstoffe in Lebensmitteln. (817.022.31), Bern, 2017. Reference Source\n\nOperation OPSON VII (2018). Commission DG Health and Food Safety, 2018. Reference Source\n\nBericht - OPSON VII: Wurde der Thunfisch «schöngefärbt»? Bundesamt für Lebensmittelsicherheit und Veterinärwesen (BLV), Bern, 2018. Reference Source\n\nAnderson RC, Wu WH: Analysis of carbon monoxide in commercially treated tuna (Thunnus spp.) and mahi-mahi (Coryphaena hippurus) by gas chromatography/mass spectrometry. J Agric Food Chem. 2005; 53(18): 7019–23. PubMed Abstract \| Publisher Full Text\n\nFrey T, Roux B, Eymann W: Detection of Carbon Monoxide-Treated Tuna by Headspace-Gas Chromatography/Mass Spectrometry. Chimia. 2006; 60: 241. Publisher Full Text\n\nNiedersächsischen Landesamt für Verbraucherschutz und Lebensmittelsicherheit (LAVES, IFF) Cuxhaven: Ein Thunfisch sieht rot - Verbrauchertäuschung durch farbstabilisierten Thunfisch. 2018. Reference Source\n\nAydin K, Aydin O: Effect of Ascorbic Acid Utilization on Cold Smoked Fish Quality (Oncorhynchus mykiss) during Process and Storage. Food Sci Technol Res. 2013; 19(5): 823–831. Publisher Full Text\n\nNiederer M: Fig03 brining experiment rawdata.csv. figshare. Dataset. 2019. http://www.doi.org/10.6084/m9.figshare.8142992.v1\n\nNiederer M: Validation with Reference samples 2017. figshare. Dataset. 2019. http://www.doi.org/10.6084/m9.figshare.8143016.v1\n\nNiederer M: Market survey 2019 rawdata. figshare. Dataset. 2019. http://www.doi.org/10.6084/m9.figshare.8143031.v1" }	[ { "id": "49814", "date": "25 Jun 2019", "name": "Ulrich Busch", "expertise": [ "Reviewer Expertise food analytics molecular biology NMR", "stable isotopes" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAddition of nitrite to tuna aiming at fresh reddish looking tuna has resulted in consumption of old tuna products with allergic reactions due to the formation of histamine. A systematic use of nitrite injections to rejuvenate tuna meat has been identified. The article \"Identification of nitrite treated tuna fish meat via the determination of nitrous oxide by head space-gas chromatography/mass spectrometry\" is well written, the method seems sufficiently valid to me. We do not know of an alternative method to prove nitrite treatment. We would be very pleased if the manuscript could be indexed, as there is no published adequate method to prove the falsification.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4820", "date": "07 Aug 2019", "name": "Markus Niederer", "role": "Author Response", "response": "We would like to thank the reviewer for his time and his important comments." } ] }, { "id": "49712", "date": "03 Jul 2019", "name": "Maurus Biedermann", "expertise": [ "Reviewer Expertise food and FCM analysis", "GC-MS", "on-line LC-GC", "comprehensive two-dimensional GCxGC" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors developed a rapid method for the detection of a nitrite treatment of tuna fish meat. It is a straight forward static headspace GC-MS method and may be most probably combined with the determination of carbon dioxide, another agent to stabilize the red color of the meat. Brining experiments with 15N labeled sodium nitrite proved the reaction path. The subject is of current importance and I suggest the indexing after minor revision.\n\nMinor comments:\nPage 3, “introduction”: Is there the possibility of combining the determination of CO and N2O, e.g. even within the same GC run? If so, please mention that option/possibility.\n\nPage 3, “Samples and sample preparation”: Please provide more details on the quantification, how was the calibration performed (one-point calibration, calibration curve, range of calibration…)?\n\nPage 4, “method development”, first and second paragraph: Within the second paragraph of the chapter it was concluded N2O to be the analyte in terms of robust chromatography. That explains the procedure of the two-step reduction described in the first paragraph. The GC temperature program was adjusted presumably based on the stronger retention of N2O. Please revise order of chapters and/or conclusions.\n\nPage 5, figure 3 text: Please mention the value of the untreated samples (below LOQ?) within the figure text.\n\nPage 5, “Validation with reference samples”: Please specify “free of N2O”; according to the supplementary data the findings ranged between 13-31 ug/kg N2O.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4819", "date": "12 Aug 2019", "name": "Markus Niederer", "role": "Author Response", "response": "We would like to thank the reviewer for his time and constructive feedback. As requested, we now mention the possibility of combining the determination of CO and N2O in the introduction. We fulfilled the request by adding the expression „one-point calibration“. Regarding the remarks on „method development“, we do not understand the line of thought: N2O not NO is the target compound. Compared to CO, N2O elutes only at higher oven temperatures. NO being a reactive intermediate is not amenable for GC analysis. With the first section explaining the fundamental differences between the detection of a CO treatment and a nitrite treatment and the second section providing details on the latter, we see no reason in revising the text. The text of fiqure 3 was not changed because the required data is already mentioned in the text of the corresponding chapter. “Validation with reference samples”: “free of N2O” is specified as lower than 30 µg/kg." } ] } ]	1	https://f1000research.com/articles/8-711
https://f1000research.com/articles/7-1466/v1	14 Sep 18	{ "type": "Software Tool Article", "title": "seqCAT: a Bioconductor R-package for variant analysis of high throughput sequencing data", "authors": [ "Erik Fasterius", "Cristina Al-Khalili Szigyarto", "Erik Fasterius" ], "abstract": "High throughput sequencing technologies are flourishing in the biological sciences, enabling unprecedented insights into e.g. genetic variation, but require extensive bioinformatic expertise for the analysis. There is thus a need for simple yet effective software that can analyse both existing and novel data, providing interpretable biological results with little bioinformatic prowess. We present seqCAT, a Bioconductor toolkit for analysing genetic variation in high throughput sequencing data. It is a highly accessible, easy-to-use and well-documented R-package that enables a wide range of researchers to analyse their own and publicly available data, providing biologically relevant conclusions and publication-ready figures. SeqCAT can provide information regarding genetic similarities between an arbitrary number of samples, validate specific variants as well as define functionally similar variant groups for further downstream analyses. Its ease of use, installation, complete data-to-conclusions functionality and the inherent flexibility of the R programming language make seqCAT a powerful tool for variant analyses compared to already existing solutions. A publicly available dataset of liver cancer-derived organoids is analysed herein using the seqCAT package, demonstrating that the organoids are genetically stable. A previously known liver cancer-related mutation is additionally shown to be present in a sample though it was not listed in the original publication. Differences between DNA- and RNA-based variant calls in this dataset are also analysed revealing a high median concordance of 97.5%.", "keywords": [ "High throughput sequencing", "whole exome sequencing", "RNA sequencing", "variant analysis", "single nucleotide variant", "R", "Bioconductor" ], "content": "Introduction\n\nHigh throughput sequencing (HTS) technologies such as genome, exome and RNA sequencing (RNA-seq) have become some of the most powerful and widely used tools in biological research worldwide, and an increasing amount of such data is being stored in online data repositories (e.g. the Gene Expression Omnibus, GEO, and the Sequence Read Archive, SRA)1–3. While decreasing experimental costs and optimised protocols enable a broad range of researchers to apply HTS to their respective scientific questions, the analysis of the resulting data is not a trivial matter, often requiring a high level of bioinformatic expertise4,5. Two examples of such software include the command line tool vcftools6 and the R-package VariantAnnotation7. There are also several software tools with a more easy-to-use graphical interface (such as the Integrative Genomics Viewer8 or the web-based Ensembl Genome Browser),9, but these are limited in their functionality. Web-based applications are limited in the amount of data that can be uploaded, and also come with the added issue of data security10. Proprietary software (such as the Ingenuity Variant Analysis)11 not only require a licence to use, but also constitute a “black box” where the underlying methods are not available for direct inspection or scrutiny.\n\nThere is thus a need for transparent, user-friendly and powerful bioinformatic tools to enable as many researchers as possible to analyse and interpret their own and publicly available HTS data. Two important aspects of such analyses is the true identity of cells analysed and comparability of both the biological samples and the data sets. Validation and evaluation of cell line authenticity, for example, is an increasingly widespread issue, as is the question of biological equivalence for any sample in general12. Here we present an open source R-package, the High Throughput Sequencing Cell Authentication Toolkit (seqCAT), which uses data from HTS experiments (whether it be of DNA or RNA origin) to investigate these matters.\n\nOne of the common outputs from HTS experiments is that of sequence variation. Single nucleotide variants (SNVs), for example, are sequence variations at the nucleotide level. Such data is the output of many variant calling programs and algorithms, which is usually stored in variant call format (VCF) files. These files are used by seqCAT in order to analyse genetic differences between samples. We have previously demonstrated the usefulness and general applicability of such analyses for both cell line authentication13 and genetic heterogeneity in public cell line datasets14. The capabilities of seqCAT include creation of SNV profiles from VCF files, comparisons of the overall genetic similarity between profiles, investigations of SNV impact distributions (i.e. variants’ predicted impact on protein function) as well as interrogations of the genotypes of previously known or user-specified variants across samples. Each individual profile can represent SNVs from a HTS experiment or from an external variant database.\n\nIn the present study, we use seqCAT to explore genetic differences within a public dataset containing both whole exome sequencing (WES) and RNA-seq data for long-term organoid cultures. We show that the organoids are genetically stable over a culture-period of several months, corroborating the original authors’ conclusions. We also demonstrate how seqCAT can be used to compare DNA- and RNA-based variant calls using the same dataset. The results highlight potential uses of variant analyses and demonstrate how seqCAT may be utilised to interrogate genetic differences at both the global and gene-specific level.\n\n\nMethods\n\nSeqCAT was developed for the Bioconductor15 repository for R-packages. It follows existing best coding practises, including a clean, modular and robust design. The basis of all seqCAT analyses are SNV profiles: collections of filtered, high-quality SNVs for any given sample. The creation of these SNV profiles is performed by filtering an input VCF file based on the available variant calling quality metrics as well as an optional sequencing depth threshold (set to ten by default)13. These profiles are saved as simple text files on the user’s hard-drive, in order to facilitate re-use and to reduce the run-time of downstream analyses. While profiles for individual samples may be created as needed by the user, several convenience-functions for working with multiple VCFs and profiles in aggregate are also available. SeqCAT can analyse VCF files with or without annotations from e.g. snpEff16 and also includes a Python-implementation of profile creation, which reduces the run-time of this step five to ten times.\n\nThe SNV profiles are subsequently read and compared to each other in a pairwise manner, yielding information on e.g. the overlap (SNVs that are present in both samples being compared), the concordance (the proportion of SNVs with identical genotypes for both samples) and the similarity score (a previously defined weighted measure of the concordance)14. Comparisons may be performed individually or in aggregate, depending on what type of analysis the user is interested in. Comparisons with external databases is also possible; seqCAT currently contains functionality to read and compare variants present in the Catalogue of somatic mutations in cancer (COSMIC) database17. Only overlapping variants are analysed by default, but non-overlaps can optionally be included as well. Examining specific chromosomes, genes or genomic regions is also possible, as are analyses of variant functionality through their predicted impact on protein-function.\n\nInstallation of both seqCAT and its dependencies is simple, and its use is described in-depth in its vignette since a major design goal of seqCAT was ease-of-use for a broad range of researchers, regardless of expertise in R. While existing data structures and objects from Bioconductor are used internally, none of these are required learning for the user; results are given as standard R-objects7,18. This makes exploration of the data as simple and easy as possible for the user. SeqCAT allows for re-analysis of already created SNV profiles, facilitating comparisons of samples across any number of datasets and includes several functions for creating publication-ready figures. All these capabilities make seqCAT a useful, simple and intuitive tool for a wide range of researchers.\n\nThe seqCAT package is designed to work with Bioconductor version 3.7 and R version 3.5.\n\n\nResults\n\nTo demonstrate the capabilities of seqCAT, we analysed a recently published dataset from Broutier et al.19. The authors created liver cancer-derived organoids for modelling disease and performed both whole exome sequencing and RNA-seq on the original tissues and the organoid cultures. We used seqCAT to analyse the raw VCF files available at GEO under accession GSE84073 (see the Supplementary Code for details and Supplementary Data 1 for the study metadata). The overall genetic similarities between tissues and organoids are clearly grouped according to their respective patient of origin, as can be seen in Figure 1A. We also investigated if this holds true for SNV profile subsets containing only coding and missense variants. The original VCF files were thus annotated using snpEff16, followed by creation, reading and sub-setting of SNV profiles. Figure 1B shows the pairwise comparisons of these variant subsets, indicating that groupings based on genetic similarities of missense variants also separate the dataset in a per-patient manner. This data covers upwards of hundreds of thousands of overlapping variants for each pairwise comparison (Table 1).\n\nPairwise comparisons of all WES SNV profiles, showing the genetic similarity of all individual samples for either no variant sub-setting (A) or sub-setting for coding variants only (B). The colour gradient is defined for ranges of the similarity score: scores between 0 and 50 are shown as white, scores between 50 and 90 as a white-to-grey gradient and, finally, a grey-to-blue gradient for 90 to 100. Samples are named according to their type: original tissues (T), established organoids (O1) and long-term cultured organoids (O2). These figures were created using the plot_heatmap seqCAT function.\n\nWe sought to investigate the genetic stability of the organoids both in terms of their transition from primary tissue to organoid culture, as well as long-term culturing. Figure 2A shows a boxplot of genetic similarities for both of these comparisons, indicating that the long-term cultures seem to be more genetically similar than the transition from tissue to organoid. This conclusion is not statistically significant, however, with p-values of 0.36 and 0.41 for all and subset variants, respectively (Supplementary Code). A larger cohort may thus be needed to fully explore the difference between tissue-to-organoid and long-term-culturing stability. The overall high genetic similarities of all the organoids are clear, however: the lowest median similarity score across all patients and all variants is 93.9 (patient CHC2), while reaching as high as 97.9 (healthy patient 1); see Table 1. The similarity scores across coding and non-subset profiles are roughly equivalent.\n\n(A) Comparisons of genetic similarities between original tissue, derived organoids and long-term cultured organoids. Results are shown for both non-subset variant comparisons and for subsets including coding variants only. The differences between T vs. O1 and O1 vs. O2 for each subset are not statistically significant (α = 0.01). (B) Analysis of previously known liver cancer SNVs as listed in the original publication, where the genotype of each individual variant is visualised by different colours. White squares indicate that no confident variant was called for that position in that particular sample. This figure was created using the plot_variant_list seqCAT function.\n\nThe original publication19 lists a number of previously known liver cancer variants (Supplementary Data 2), which we analysed with seqCAT. This analysis reveals that some of the known variants are present in the organoids but absent in their corresponding tissue (Figure 2B). SeqCAT indicates that these specific variants would need to be investigated further, which the original authors have done in most cases. However, it revealed that the GPRIN1 variant is present in the CC1 samples, something not mentioned in the original publication.\n\nAnnotations with snpEff include variant impacts, which are the predicted effects on protein functions and range from HIGH, MODERATE, LOW through MODIFIER, in decreasing order of importance. An example of a HIGH impact is a variant leading to protein truncation, while a MODIFIER variants is predicted to a little to no effect on their resulting protein (such as intronic variants). SeqCAT can summarise and visualise these impacts across profile comparisons. Figure 3 shows the impact distributions of matching and mismatching variants for an aggregation of all comparisons between samples in the tissue-to-organoid transition as well as through the long-term culturing process. There is a higher proportion of mismatching MODIFIER variants, and there are only a limited number of mismatching HIGH variants.\n\nDistribution of variant impacts for the an aggregate of all pairwise comparisons between tissue and early organoid cultures (A), and early versus late organoid cultures (B). Matching variants (i.e. variants with identical genotypes for both samples being compared) are dark blue, while mismatching variants are a lighter shade of blue. These figures were created using the plot_impacts seqCAT function.\n\nIn order to investigate if any of these mismatching variants are biologically relevant, we performed GO (Gene Ontology) and KEGG (Kyoto Encyclopedia of Genes and Genomes) enrichment using DAVID20 on genes affected by mismatching variants in the HIGH and MODERATE impact categories. While no terms were significantly enriched for the tissue-to-organoid transition (α = 0.01), three olfactory-related terms and one related to protein de-ubiquitination were significantly enriched for long-term culturing comparisons (see Supplementary Data 3 and Supplementary Data 4).\n\nIn summary, these results corroborate the original authors’ conclusion that the organoids are accurate and genetically stable in vitro models of liver cancer and demonstrate how seqCAT can be used to analyse genetic variation in HTS data.\n\nThe Broutier dataset contains not only WES data but also RNA-seq data on the same samples, enabling comparison of RNA-seq data to the already performed WES analyses. We thus downloaded the publicly available raw FASTQ files, performed read alignment with the 2-pass mode of STAR21, variant calling using GATK22 and annotation using snpEff16, as previously described13. We subsequently used seqCAT to create SNV profiles for each RNA-seq sample and performed pairwise comparisons across all WES and RNA-seq SNV profiles. This resulted in a grouping with high similarities between WES and RNA-seq samples for the same patient (Figure 4).\n\nThe colour gradient is the same one used for Figure 1: scores between 0 and 50 are white, scores between 50 and 90 are shown with a white-to-grey gradient, and a grey-to-blue gradient for scores between 90 and 100. This figure was created using the plot_heatmap seqCAT function.\n\nThere are several previously published studies that show discrepancy between DNA and RNA variants with varying extent and proposed causes23,24. In order to quantify the differences between DNA- and RNA-based variants in the organoid dataset, the median concordance for all same-sample comparisons was calculated to be 97.5%; the concordance was used in lieu of the similarity score in order to increase comparability with previously published results. This was also performed for sample type-specific comparisons, where the concordance for tissue versus tissue comparisons was 96.5% and 97.7% for organoid versus organoid. Per-patient (e.g. CC1 vs. CC1) calculations were also performed, shown in Table 2. The minimum per-patient concordance was 94.8% and the maximum 98.9%, while the minimum for any individual comparison was 81.1% and a maximum of 99.0% (see the Supplementary Code for the calculations). The minimum value of 81.1% (tissue versus tissue for patient CC1) is the only DNA/RNA comparison with a concordance lower than 90%. These concordances are generally higher than the 80 to 90% that have previously been shown24.\n\nIn summary, results from seqCAT demonstrate an overall high level of concordance between DNA and RNA variant calls, but highlight that there is some variation between sample types and patients.\n\n\nDiscussion\n\nHTS experiments are becoming increasingly more common and the need for simple and powerful bioinformatic software is as great as ever. Analyses of genetic variation through e.g. SNVs represents a common endeavour for many scientific studies, but the methods and data analysis pipelines used vary. In this study we present seqCAT, an easy-to-use and well-documented Bioconductor15 R-package that performs variant analyses of HTS data. The capabilities of seqCAT include the creation of SNV profiles (including a five to ten times faster implementation in Python), comparisons of global genetic similarities for all variants common between samples and analyses of single variants or genes of special interest. While the seqCAT package itself is new, the underlying theory and general methodology have previously been used for investigations into cell line authenticity13 and genetic heterogeneity in public cell line datasets14.\n\nSeqCAT may be used to analyse both novel sequencing data as well as publicly available data in repositories (such as the GEO)1, but may also be utilised to define genetic profiles for any sample of interest. Such profiles are of great interest for researchers using model systems (such as cell lines or organoids), as it allows for a clear definition of the genetic background of the model itself. This could then be referred back to at a later time, to make sure that genetic drift (that obscure interpretation of biological results) has not occurred. SeqCAT is both easy to install and to use, and includes in-depth documentation on its functionality and underlying theory.\n\nIn the present study, we have used seqCAT to analyse a publicly available dataset containing WES and RNA-seq data from organoid cultures and their tissues-of-origin19. The global analysis of WES SNVs demonstrate the overall high genetic similarities between the organoids and their respective tissues, with equivalent results for comparisons covering all variants or only missense variants. The seqCAT-analysis of known variants indicate that a GPRIN1 variant is present for the CC1 patient; this variant is only listed as present in a CHC-type patient in the original study. Given the importance of these previously known variants it is likely that the GPRIN1 mutation may be of significance not only for the originally listed CHC1 patient, but also for CC1 patients. The results presented herein corroborate the original authors’ conclusions that organoids are genetically stable over time, but higher level of genetic similarity between early and long-term cultured organoids as compared to the tissue-to-organoid transition is statistically non-significant.\n\nThe analyses of genes affected by mismatching HIGH and MODERATE impact variants show that none of the differences between tissue and initial organoid cultures are significantly enriched for specific biological functions, indicating that these differences likely are random. The transition from primary tissue to organoid can thus be viewed as a highly stable transition, especially given the high overall similarity previously discussed. The long-term culturing results do, however, present four significantly enriched terms. Three of these are related to ectopic expression of olfactory receptors, which have previously been shown to be present in both healthy and cancerous tissues25,26. The single GO-term related to protein de-ubiquitination may be important for studies investigating ubiquitination in liver cancer. Both of these points should thus be accounted for when performing a study with these organoids. The overall results yielded by the seqCAT-analyses corroborate the conclusions from the original study, i.e. that these organoids are genetically stable and suitable models for studying liver cancer.\n\nThere have been several studies comparing variant calls from DNA and RNA of the same samples, but they have come to differing conclusions as to both the extent and causes of the DNA/RNA discrepancies. Li et al. performed both DNA/RNA-seq across 27 individuals in addition to analyses of protein expression using mass spectrometry, where peptides corresponding to variants found in both DNA and RNA were present23. They argue that their results indicate biological significance of RNA variants, given that they are translated to proteins, and that the differences between DNA and RNA variants can be biologically meaningful. Indeed, there have been several studies analysing RNA-seq variants that yielded novel biological insights, demonstrating the utility of such endeavours27–31. A study by Guo et al. analysed DNA/RNA-seq data for 10 breast cancer patients from the TCGA and calculated DNA/RNA concordances to range between from 80 to 90%24. They argue that these differences are mostly technical rather than biological.\n\nThe results of the present study indicate that the extent of DNA/RNA differences may not be as large as previously shown: the median concordance for DNA/RNA pairs was 97.5% overall, with a range of 90 to 99% (plus a single comparison with 81.1%), while Guo et al. reported a range of 80 to 90% concordance. Both studies thus find a discrepancy between DNA- and RNA-based variant calls, but disagree on its extent. The RNA-seq pipeline utilised in this study is based on the current best practices of GATK, which uses the STAR software for read alignment that has proven to be highly accurate for RNA-seq data21,32. The latest assembly of the human genome (GRCh38) was also used, as the choice of assembly has been highlighted as an important parameter that can yield higher accuracy24. Guo et al. used an earlier assembly from 2009 (GRch37), which might partly explain the discrepancy between the results. While technical issues will always exist even for DNA/DNA or RNA/RNA comparisons, the results of the present study may represent a closer estimate of the biological relevance of DNA/RNA differences first noted by Li et al.\n\nIt is clear is that there is a discrepancy between DNA- and RNA-based variant calls, but the exact extent of this difference remains to be determined, as well as whether it is a consequence of technical artefacts or biological variation. A full evaluation of these matters likely require a larger study than what has previously been attempted, including using the latest technologies as well as protein-level validation. The analyses performed herein demonstrate how seqCAT may be utilised as a part of such an endeavour.\n\n\nConclusions\n\nThe seqCAT Bioconductor R-package provides an effective and easy-to-use toolkit for analysing HTS variant data, enabling researchers to investigate genetic differences and potential variation within and between their samples or publicly available data from other laboratories. Little R expertise is required to use seqCAT, and its use is extensively documented. We have used seqCAT to analyse genetic variation in a publicly available dataset of liver cancer organoids, corroborating the conclusions drawn by its original authors, as well as demonstrate high levels of DNA/RNA SNV concordance in this dataset. These results serve as a case study in how to utilise the capabilities of seqCAT, which make it a valuable and intuitive tool for a wide range of researchers.\n\n\nSoftware and data availability\n\nSoftware is available from: https://bioconductor.org/packages/release/bioc/html/seqCAT.html\n\nSource code available from: https://github.com/fasterius/seqCAT\n\nArchived source code as at time of publication: https://doi.org/10.5281/zenodo.140402733\n\nSoftware license: MIT\n\nThe data used in this article is publicly available at the GEO through the accession number GSE84073.", "appendix": "Grant information\n\nThis work was supported by the European Community 7th Framework Program under grant agreement no. 278 568 “PRIMES”.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe would like to acknowledge support from Science for Life Laboratory (SciLifeLab), the National Genomics Infrastructure (NGI) and Uppmax for providing assistance in computational infrastructure.\n\n\nSupplementary material\n\nSupplementary Code: A RMarkdown document for reproducing the analyses and figures of the study using the seqCAT package.\n\nClick here to access the data.\n\nSupplementary Data 1: Metadata for the Broutier et al. study19.\n\nClick here to access the data.\n\nSupplementary Data 2: List of the previously known SNVs used in the Broutier et al. study19.\n\nClick here to access the data.\n\nSupplementary Data 3: Full results of the enrichment analysis of tissue versus established organoids.\n\nClick here to access the data.\n\nSupplementary Data 4: Full results of the enrichment analysis of established organoids versus long-term cultured organoids.\n\nClick here to access the data.\n\n\nReferences\n\nEdgar R, Domrachev RM, Lash AE: Gene Expression Omnibus: NCBI gene expression and hybridization array data repository. Nucleic Acids Res. 2002; 30(1): 207–210. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhu Y, Stephens RM, Meltzer PS, et al.: SRAdb: query and use public next-generation sequencing data from within R. BMC Bioinformatics. 2013; 14: 19. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHeather JM, Chain B: The sequence of sequencers: The history of sequencing DNA. Genomics. 2016; 107(1): 1–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSboner A, Mu XJ, Greenbaum D, et al.: The real cost of sequencing: higher than you think! Genome Biol. 2011; 12(8): 125. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMuir P, Li S, Lou S, et al.: The real cost of sequencing: scaling computation to keep pace with data generation. Genome Biol. 2016; 17: 53. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDanecek P, Auton A, Abecasis G, et al.: The variant call format and VCFtools. Bioinformatics. 2011; 27(15): 2156–2158. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nObenchain V, Lawrence M, Carey V, et al.: VariantAnnotation: a Bioconductor package for exploration and annotation of genetic variants. Bioinformatics. 2014; 30(14): 2076–2078. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRobinson JT, Thorvaldsdóttir H, Winckler W, et al.: Integrative genomics viewer. Nat Biotechnol. 2011; 29(1): 24–26. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZerbino DR, Achuthan P, Akanni W, et al.: Ensembl 2018. Nucleic Acids Res. 2018; 46(D1): D754–D761. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPabinger S, Dander A, Fischer M, et al.: A survey of tools for variant analysis of next-generation genome sequencing data. Brief Bioinform. 2014; 15(2): 256–278. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nQIAGEN: Ingenuity Variant Analysis. 2018; accessed 2018-05-30. Reference Source\n\nCapes-Davis A, Neve RM: Authentication: A Standard Problem or a Problem of Standards? PLoS Biol. 2016; 14(6): e1002477. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFasterius E, Raso C, Kennedy S, et al.: A novel RNA sequencing data analysis method for cell line authentication. PLoS One. 2017; 12(2): e0171435. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFasterius E, Al-Khalili Szigyarto C: Analysis of public RNA-sequencing data reveals biological consequences of genetic heterogeneity in cell line populations. Sci Rep. 2018; 8(1): 11226. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHuber W, Carey VJ, Gentleman R, et al.: Orchestrating high-throughput genomic analysis with Bioconductor. Nat Methods. 2015; 12(2): 115–121. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCingolani P, Platts A, Wang le L, et al.: A program for annotating and predicting the effects of single nucleotide polymorphisms, SnpEff: SNPs in the genome of Drosophila melanogaster strain w1118; iso-2; iso-3. Fly (Austin). 2012; 6(2): 80–92. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nForbes SA, Beare D, Gunasekaran P, et al.: COSMIC: exploring the world’s knowledge of somatic mutations in human cancer. Nucleic Acids Res. 2015; 43(Database issue): D805–11. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLawrence M, Huber W, Pagès H, et al.: Software for computing and annotating genomic ranges. PLoS Comput Biol. 2013; 9(8): e1003118. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBroutier L, Mastrogiovanni G, Verstegen MM, et al.: Human primary liver cancer-derived organoid cultures for disease modeling and drug screening. Nat Med. 2017; 23(12): 1424–1435. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHuang da W, Sherman BT, Lempicki RA: Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc. 2008; 4(1): 44–57. PubMed Abstract \| Publisher Full Text\n\nDobin A, Davis CA, Schlesinger F, et al.: STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013; 29(1): 15–21. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcKenna A, Hanna M, Banks E, et al.: The Genome Analysis Toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res. 2010; 20(9): 1297–1303. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLi M, Wang IX, Li Y, et al.: Widespread RNA and DNA sequence differences in the human transcriptome. Science. 2011; 333(6038): 53–58. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGuo Y, Zhao S, Sheng Q, et al.: The discrepancy among single nucleotide variants detected by DNA and RNA high throughput sequencing data. BMC Genomics. 2017; 18(Suppl 6): 690. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFlegel C, Manteniotis S, Osthold S, et al.: Expression profile of ectopic olfactory receptors determined by deep sequencing. PLoS One. 2013; 8(2): e55368. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAbaffy T: Human olfactory receptors expression and their role in non-olfactory tissues-a mini-review. J Pharmacogenomics Pharmacoproteomics. 2015. 6: 152. Publisher Full Text\n\nMiller AC, Obholzer ND, Shah AN, et al.: RNA-seq-based mapping and candidate identification of mutations from forward genetic screens. Genome Res. 2013; 23(4): 679–686. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPiskol R, Ramaswami G, Li JB: Reliable identification of genomic variants from RNA-seq data. Am J Hum Genet. 2013; 93(4): 641–651. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLee MC, Lopez-Diaz FJ, Khan SY, et al.: Single-cell analyses of transcriptional heterogeneity during drug tolerance transition in cancer cells by RNA sequencing. Proc Natl Acad Sci U S A. 2014; 111(44): E4726–E4735. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDeelen P, Zhernakova DV, de Haan M, et al.: Calling genotypes from public RNA-sequencing data enables identification of genetic variants that affect gene-expression levels. Genome Med. 2015; 7(1): 30. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKang HM, Subramaniam M, Targ S, et al.: Multiplexed droplet single-cell RNA-sequencing using natural genetic variation. Nat Biotechnol. 2018; 36(1): 89–94. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEngström PG, Steijger T, Sipos B, et al.: Systematic evaluation of spliced alignment programs for RNA-seq data. Nat Methods. 2013; 10(12): 1185–1191. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFasterius E, vobencha, hpages: fasterius/seqCAT: seqCAT version 1.2.1 (Version 1.2.1). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1404027" }	[ { "id": "46167", "date": "09 Apr 2019", "name": "Matej Lexa", "expertise": [ "Reviewer Expertise Bioinformatics" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript describes a novel computational tool for genotype analysis and comparison called seqCAT.\nThe tool has been created as a package for R/Bioconductor and has already been accepted into the Bioconductor repository. I was able to install it and follow the example code in the package as well as study its use in the manuscript. In all my tests the package and its functions worked as designed/described in the accompanying materials.\nAlthough the code is fully functional, both the code and the submitted manuscript leave much to be deserved. The most important issues in this respect are i) the absense of critical comparison with existing tools, ii) better description of some of the available functionality and last but not least, iii) better integration into the existing data and code structure.\nAs far as other tools are regarded, the authors cite the need for a tool like seqCAT by referring to vcftools, VariantAnnotation R package, IGV and Ensembl Genome Browser and some proprietary software. However, today there are dozens of tools that may come close to the functionality presented here and deserve to be mentioned and compared critically. Just a quick browsing of several sources yielded software, such as adegenet (https://cran.r-project.org/web/packages/adegenet/index.html), anvi'o (http://merenlab.org/2015/07/20/analyzing-variability/), SomaticSniper (http://gmt.genome.wustl.edu/packages/somatic-sniper/), PhyloSNP (https://hive.biochemistry.gwu.edu/dna.cgi?cmd=phylosnp), GATK or BEDOPS that has a vcf2bed function (https://bedops.readthedocs.io/en/latest/content/reference/file-management/conversion/vcf2bed.html) that can lead to comparison based on interval sets. Would PLINK and its SNP profiling abilities be powerful enough (http://zzz.bwh.harvard.edu/plink/profile.shtml)? Are methods typically used for small and medium-sized SNP samples, such as the MATLAB code here (https://jamanetwork.com/journals/jamaoncology/fullarticle/2598491) different from methods that must be applied to whole-genome data? I don't know the answers to some of these questions but I feel the authors should look wider to show the advantages of seqCAT, if any. One advantage, also mentioned by the authors is simplicity of use. However, it should be clear what the trade-offs are. The manuscript mentions SNVs are filtered based on quality and other criteria but doesn't give enough details about what is happening under the hood. The software is open source, however the manuscript should lay out basic principles of data manipulation done by their package in plain English. Also, reading a profile into a package and comparing it to others create different GenomicRanges/data frame data objects in R that should also be described briefly. Loosely connected to the data frame data structures mentioned above, I see the way seqCAT manipulates data as a weak point. First of all, it calculate profiles and saves them into a file, effectively outside R, only to read the files in the next step. It would seem much more natural, to use some internal data structure, maybe even the same data frame created later, to keep the data in R and offer appropriate writing/reading/conversion functions to create files outside R. As for conversions, data formats for some of the data calculated by seqCAT already exist and would make the software much more powerful, if the users could write to them (or even read from them). Although the profiles can be exported into BED/GFF3 with some third party libraries (e.g. rtracklayer), perhaps it would be useful to go to BAM/SAM, back to VCF after some manipulation (right now only filtration, presumably), or hapmap and others for transfer of data into other software (e.g. PLINK, VarDict)?\n\nIs the rationale for developing the new software tool clearly explained? Partly\n\nIs the description of the software tool technically sound? Partly\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4762", "date": "12 Aug 2019", "name": "Cristina Al-Khalili Szigyarto", "role": "Author Response", "response": "We are thankful for the feedback provided by the reviewer, and have made several changes to both the seqCAT package itself, its documentation and the manuscript. The latter’s introductory section has been extended with a more thorough exploration of existing tools and how seqCAT differs from them. The description of the various filtering criteria has also been extended, along with a new section covering the same in the package vignette. The structure and use of data objects have been streamlined to only utilise data frames at the user-level, to further facilitate easy-of-use and consistency (version 1.7.1). The creation of SNV profiles is now performed within R, with no saving of the final profile to the hard disk; a separate function for this has been added as an option for users that still desire the old functionality (1.7.1). Reading and writing in several other file formats has also been added (1.7.2)." } ] }, { "id": "49270", "date": "10 Jun 2019", "name": "Jean Fan", "expertise": [ "Reviewer Expertise bioinformatics", "software development", "RNA variant calling", "cancer biology" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nOverview\nFasterius and Szigyarto present seqCAT, an R-package for single nucleotide variant analysis (downstream of alignment and variant calling). The authors apply seqCAT to characterize the genetic concordance between DNA and RNA samples as well as cancer samples and derived organized.\nWhile high throughput sequencing technologies are indeed producing large amounts of big data demanding novel computational tools and software for efficient processing and analysis, it is unclear why the analysis presented in this manuscript demanded a new software rather than applying functionalities already existing in packages such as VariantAnnotation and GenomicRanges, both which the presented software builds on. The software, while comparably easy to use to other Bioconductor packages, is not sufficiently well documented, particularly with respect to the variant filtering step. Likewise, the presented analysis does not highlight the utility or necessity of the developed software. Generally, the rationale for the development of the software is unclear, particularly given its redundancy with existing software.\n\nTherefore, this manuscript and software in its current form is not of high enough standard to warrant indexing. I hope the following comments will be useful for the authors.\n\nMajor comments (software)\nBeing that this is a software tool article, I have compared my typical VCF analysis approach to seqCAT's pipeline and have the following set of major and minor comments:\n1. Regarding the create_profile() function\n\n```{r} # Start with same VCF file library(\"seqCAT\") vcf <- system.file(\"extdata\", \"example.vcf.gz\", package = \"seqCAT\")\n# How I would typically parse a VCF file library(VariantAnnotation) data <- readVcfAsVRanges(vcf) vi <- sampleNames(data) == \"HCT116\" hct.norm <- data[vi,] class(hct.norm) length(hct.norm)\n# How seqCAT does it setwd(\"~/Desktop\") # note a text file is written so I need to change my working directory to somewhere that I have write permission create_profile(vcf, \"HCT116\", \"hct116.profile.txt\", filter=FALSE) hct116 <- read_profile(\"hct116.profile.txt\", \"HCT116\") # now I have to read the written file it back in class(hct116) length(hct116) ```\nMy approach reads in a VCF file as VRanges and filters for the variants annotated as HCT116 for the sample using functionalities available in the VariantAnnotation package, ending up with 12055 variants. In comparison, with seqCAT, I only get 1210 variants, despite setting the filter parameter to FALSE. The documentation is not sufficiently clear for me to understand the cause of the difference.\n\nI can speculate that perhaps seqCAT is restricting to the single-nucleotide variants as opposed to larger indels. So I can test this:\n```{r} # limit to single nucleotide variants vi <- width(ranges(hct.norm)) == 1 hct.norm <- hct.norm[vi,] length(hct.norm) ```\nHowever, this still leaves me with 10773 variants, magnitudes more than the 1210 variants identified by seqCAT. The filtering criteria used by seqCAT are not well documented.\n\n2. Regarding the compare_profiles() function\n```{r} # Begin with same set of variants create_profile(vcf, \"HCT116\", \"hct116.profile.txt\", filter=TRUE) hct116 <- read_profile(\"hct116.profile.txt\", \"HCT116\") create_profile(vcf, \"RKO\", \"rko.profile.txt\", filter=TRUE) rko <- read_profile(\"rko.profile.txt\", \"RKO\")\n# How I would typically compare two VCF files # to count number of shared variants length(intersect(hct116, rko)) # or if I want to keep the detailed info foo <- unique(hct116[hct116 %in% rko,])\n# How seqCAT does it hct116_rko <- compare_profiles(hct116, rko) dim(hct116_rko) class(hct116_rko) ```\nIn both cases, I end up with 282 shared variants between HCT116 and RKO, so it is unclear why the compare_profiles() function is necessary given existing intersect() functionalities already available. Furthermore, note that the resulting output of seqCAT's compare_profiles() function is a dataframe object, even though the inputs are both GRanges objects, whereas my approach maintains a GRanges data structure. It is unclear why seqCAT casts the GRanges input into dataframes rather than maintaining a consistent data structure.\n\nMinor comments (software)\n1. The authors note that seqCAT \"follows existing best coding practises, including a clean, modular and robust design.\" Please elaborate on what these existing best coding practices are or cite the referenced best coding practices if possible. Also note the misspelling.\n\n2. The seqCAT software provides additional functionalities \"to read and compare variants present in the Catalogue of somatic mutations in cancer (COSMIC) database\" as noted in the manuscript. However, these functionalities do not appear used in the analyses presented in the manuscript. Out of curiosity, are any of the genetic differences between primary cancer samples and derived organoids found in COSMIC? Likewise, what is the genetic similarity measurement when subsetted to variants present in COSMIC?\nMajor comments (biology)\nWhile this is a software tool article, the authors apply the software to analyze biological data and propose a number of biological conclusions, for which I have the following set of major and minor comments:\n1. The authors propose that \"long-term cultures seem to be more genetically similar than the transition from tissue to organoid.\" However, the analysis performed to support this claim was limited to single nucleotide variants. So it remains unclear whether larger-scale genetic alterations are present and whether long-term cultures are also genetically more similar than the transition from tissue to organoid in terms of these other non-single-nucleotide genetic alternations.\n\n2. The authors note that \"there are only a limited number of mismatching HIGH variants\" when comparing tissue and early organoid cultures. However, closer inspection finds that the proportion of mismatching HIGH variants (0.2%) is quite comparable to the proportion of matching HIGH variants (0.3%). Based on this result, the null hypothesis that both matching and mismatching variants come from the same underlying impact distribution cannot be rejected. It is inaccurate to imply that mismatching variants are depleted in the HIGH impact category simply because there are so few of them as the total number of mismatching variants is also fewer than matching variants.\n\n3. Why was the GPRIN1 variant missed in the original publication and found by seqCAT? It is unclear whether this discrepancy is the result of alignment, variant calling, and other upstream differences or if it is the result of using seqCAT. How do we know this mutation is not a false positive/sequencing error? What is GPRIN1? Is it an important oncogene? Is the mutation present in COSMIC? \"Given the importance of these previously known variants it is likely that the GPRIN1 mutation may be of significance\" is not sufficiently justified.\n\nMinor comments (biology)\n1. The authors find that \"Differences between DNA- and RNA-based variant calls in this dataset are also analysed revealing a high median concordance of 97.5%.\" This is still not 100% So are the differences enriched for putative RNA editing? Or do they occur at SNVs with lower quality scores? What explains the difference?\n2. The conclusion \"organoids are accurate...in vitro models of liver cancer\" is not adequately supported without a more thorough transcriptomics comparison of the organoids and primary cancer tissue. While such a study is beyond the scope of this manuscript, the authors should refrain from drawing inadequately supported conclusions.\n\n3. The authors find that their DNA and RNA variant calls exhibit ~97.5% median concordance compared to previous 80 to 90% estimates, speculating that this increased concordance is solely the result of improved alignment, variant calling best practices, and using the latest human genome assembly, rather than due to use of seqCAT. It is unclear whether or not sequencing technology differences, or even biological differences between breast cancer and liver cancer are also contributing factors.\n\nIs the rationale for developing the new software tool clearly explained? No\n\nIs the description of the software tool technically sound? Partly\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? No", "responses": [ { "c_id": "4763", "date": "12 Aug 2019", "name": "Cristina Al-Khalili Szigyarto", "role": "Author Response", "response": "We are grateful for the constructive and thorough criticism presented by the reviewer, and have implemented several alternations to both the seqCAT package itself (version 1.7.1 and forwards), its documentation and the manuscript. Ease-of-use is one of the main features of seqCAT, as the reviewer has already pointed out, along with a complete data-to-figures workflow. The manuscript’s introduction has been extended to more fully address these points and motivate the rationale behind seqCAT’s creation. The discrepancy between the process used by the reviewer and seqCAT can indeed be attributed to the additional filtering that seqCAT performs. This filtering includes variant caller-specific filtering, minimum variant depth, mitochondrial variants, non-standard chromosome, unique variants at the gene- or position-level and variants with missing genotypes. We agree with the reviewer that this should be made clearer in both the manuscript and the documentation, and have thus made appropriate changes (version 1.7.2). As already pointed out by the reviewer and as stated in the manuscript, seqCAT only works with SNV profiles, i.e. single nucleotide variants. The reason the `compare_profiles` function does not use code similar to that shown in the reviewer’s example is that such a procedure, while indeed yielding the same overlaps, loses the sample-specific metadata. Looking at the number of metadata-columns for the example comparison there is a difference of 11 (18 using the reviewer’s code, 29 for seqCAT). These sample-specific metadata is used both in the `compare_profiles` itself (such as for comparing sample genotypes) as well as in downstream analyses (such as when plotting variant grids or impact distributions). SNV profile creation is now performed within R and the mandatory storage of profiles to disk has been removed; this can now be performed with a separate function for users that still require it. Inconsistencies in seqCAT’s data structure has been addressed: now only data frames are presented to the user, in order to facilitate simplicity and ease-of-use. Several updates and extensions to the manuscript has also been added, addressing the biology-related comments raised by the reviewer." } ] } ]	1	https://f1000research.com/articles/7-1466
https://f1000research.com/articles/8-1400/v1	09 Aug 19	{ "type": "Research Article", "title": "Non-AIDS defining malignancies in the combination ART era: immunological and socio-behavioral risk factors", "authors": [ "Yann Ruffieux", "Frédérique Chammartin", "Anita Feller", "Kurt Schmidlin", "Sabine Rohrmann", "Katharina Staehelin", "Christine Bouchardy", "Heiner C. Bucher", "Barbara Hasse", "Matthias Cavassini", "Alexandra Calmy", "Gilles Wandeler", "Alexandra Scherrer", "Julia Bohlius", "Matthias Egger", "Yann Ruffieux", "Frédérique Chammartin", "Anita Feller", "Kurt Schmidlin", "Sabine Rohrmann", "Katharina Staehelin", "Christine Bouchardy", "Heiner C. Bucher", "Barbara Hasse", "Matthias Cavassini", "Alexandra Calmy", "Gilles Wandeler", "Alexandra Scherrer", "Julia Bohlius" ], "abstract": "Background: Since the advent of combination antiretroviral therapy (cART), non-AIDS defining malignancies (NADM) have become increasingly important. We examined risk factors for NADM, including immunological, virological and socio-behavioral characteristics. Methods: We linked the Swiss HIV Cohort Study (SHCS) with cancer registries to identify incident cancers between 1996 and 2012. We analyzed four common NADM: anal, lung, prostate, and liver cancer. We calculated standardized incidence ratios (SIRs) and assessed the effect of time-updated CD4 and CD8 count, CD4/CD8 ratio, and HIV viral load (copies/ml) in Cox regression models. We lagged time-dependent variables for 12, 24, and 36 months and captured cumulative exposures using simple moving averages (SMA). In multivariable models, we also considered HIV transmission group, smoking, and chronic hepatitis B or C infection as potential predictors of NADM incidence. Results. Between 1996 and 2012, 563 HIV-infected individuals developed NADM, including 70 anal, 49 lung, 44 prostate, and 36 liver cancers. Compared with the general population, the SHCS exhibited higher rates of anal (SIR 76.1, 95% Confidence interval (CI) 60.2-96.2), lung (SIR 1.98, 1.50-2.62), and liver cancer (SIR 7.28, 5.25-10.1) but similar rates of prostate cancer (SIR 1.03, 0.76-1.38). Anal cancer was associated with low CD4 cell count, high CD8 cell count, men who have sex with men, and smoking. For lung cancer, the CD8 cell count was the only significant predictor identified among the immunological and virological factors. CD4 cell count, and chronic hepatitis B and C infection were predictive of liver cancer incidence. We found no evidence of any of the immunological factors being associated with prostate cancer. Conclusions: The importance of immunodeficiency (indexed by CD4 count) and immune senescence (indexed by CD8 count) differs across NADM. Immunodeficiency was an important risk factor for anal and liver cancer whereas immune senescence was associated with lung cancer and anal cancer.", "keywords": [ "Non-AIDS defining cancers", "cohort studies", "coinfection", "hepatitis" ], "content": "Introduction\n\nIn high-income countries, life expectancy of people living with HIV (PLWHIV) on combination antiretroviral therapy (cART) is approaching that of the general population1. For example, in Switzerland 2006 to 2013, PLWHIV with higher education had an estimated life expectancy similar to individuals from the general population with compulsory education2. Similarly, in Canada 2008 to 2012, life expectancy at age 20 years in HIV-positive people was 89% of the life expectancy in the general population3. In line with the increasing life expectancy, morbidity and mortality of PLWHIV on cART are no longer dominated by AIDS-defining events, but also affected by cardiovascular events and non-AIDS defining malignancies (NADM)1–5.\n\nAn analysis of causes of death in the Data collection on Adverse events of Anti-HIV Drugs (a collaboration of eleven HIV cohort studies from Europe, the USA and Australia), found that mortality decreased from 1999 to 2011 for most causes of death, including AIDS-related, liver-related, cardiovascular and other or unknown causes4. However, in this collaborative study4 and in other cohorts5–7, mortality from NADM remained stable over time. This type of cancer is now the most common cause of non-AIDS deaths in PLWHIV in high-income countries8. Furthermore, compared to the general population, the incidence of NADM is increased in PLWHIV, with a standardized incidence ratio for all NADM combined ranging from 1.6 to 2.8 across different studies9.\n\nSeveral factors contribute to the higher risk of NADM among HIV-positive individuals, including prolonged exposure to immunodeficiency, a higher prevalence of smoking, and infections with oncogenic viruses such as human papilloma virus (HPV) or hepatitis B (HBV) or hepatitis C (HCV) viruses10–12. Direct pro-oncogenic effects of HIV, inflammation, and coagulation pathways may also play a role13,14. More recently, accelerated immune senescence, as evidenced by low ratios of CD4 to CD8 positive T cells (CD4/CD8 ratio), has been identified as a potential risk factor for NADM15,16. For example, in the US Veterans Aging Cohort Study, the CD4/CD8 ratio, but not the CD4 cell count, was an independent risk factor for lung cancer15.\n\nWe linked the database of the Swiss HIV Cohort Study (SHCS) with cantonal cancer registries to examine the relative importance of different risk factors for NADM, including immunological, virological and socio-behavioral factors.\n\n\nMethods\n\nThe SHCS is an ongoing observational study that promotes clinical and public health research on HIV in Switzerland17,18. The SHCS is a collaboration between various hospitals, clinics, laboratories, and private practices across the country. Prospective enrollment started in 1988, and involves PLWHIV aged 16 years or above. Detailed socio-demographic and behavioral parameters for each patient are collected at enrollment. Laboratory results such as CD4+ and CD8+ cell count, HIV-1 RNA viral load or indicators of exposure to hepatitis B or C virus are recorded at enrollment and during follow-up visits. Detailed information on ART and other treatments, and of AIDS defining events are also collected at regular intervals. While NADM are recorded in the study, their ascertainment is incomplete.\n\nIncident cancers in the SHCS were identified through linkage of SHCS records with the records of the Basel, Geneva, and Zurich cantonal cancer registries. At the time of the linkage, the data covered the period up to 2012 for Geneva and Zurich, and 2011 for Basel. Only SHCS participants registered in the areas covered by the three aforementioned cancer registries were included, representing about 65% of all participants. The linkage was done using privacy-preserving probabilistic record linkage19,20. Briefly, we employed a three-party protocol, with the SHCS and cancer registries collecting confidential individual-level patient data and the Institute of Social and Preventive Medicine at University of Bern serving as the independent linkage center. The linkage center supplied purpose-written encryption software21 that was used at SHCS centers and cancer registries to irreversibly encrypt personal identifying data (names, date of birth) using Bloom filters and calculate the similarity coefficients, which were then used to identify matching records at ISPM.\n\nWe supplemented the cancer cases reported in the SHCS by the cases identified through data linkage. We analyzed NADM classified according to the 10th revision of the International Classification of Diseases (ICD-10) that were diagnosed in about 40 or more patients: malignant neoplasm of anus and anal canal (C21), malignant neoplasm of liver and intrahepatic bile ducts (C22, excludes malignant neoplasm of other and unspecified parts of biliary tract), malignant neoplasm of trachea, bronchus and lung (C33–34), and prostate cancer (C61). We did not include Hodgkin lymphoma because most cases had been diagnosed in the early years of cART, when CD8 counts and viral load measurements were often missing.\n\nWe calculated standardized incidence ratios (SIRs) for each identified NADM by comparing the observed number of cancers among people with HIV with the expected number of occurrences, based on age-, sex- and period-specific rates for the Swiss population provided by the National Institute for Cancer Epidemiology and Registration (indirect standardization)22. We produced 95% confidence intervals (CI) for the SIRs assuming a Poisson distribution23.\n\nWe assessed the effect of baseline and time-updated exposure variables in Cox regression models. Time-updated variables included CD4 cell count (cells/µL), CD8 cell count (cells/µL), CD4/CD8 ratio and HIV viral load (copies/mL). We updated the immunological and virological factors at the start of each month. Months with no CD4 cell count, CD8 cell count, or HIV viral load were interpolated by carrying forward or carrying back the closest measured value within 12 months, provided there was no change in cART and no diagnosis of cancer in the interval. Time-updated bacterial pneumonia was additionally considered for lung cancer: history of bacterial pneumonia changed from no to yes when pneumonia was diagnosed. Other variables included age, likely HIV transmission group (men having sex with men [MSM], intravenous drug users [IDU], other men, other women), current smoking at enrollment (yes, no), highest completed level of education (compulsory, secondary, tertiary), chronic hepatitis B and exposure to hepatitis C (yes or no, as a time-updated variable), and calendar period (time-updated; before 2002, 2002–2007, 2008 and later). Patients were defined as having chronic hepatitis B if they had previously been tested positive for HBs-antigen or HBV-DNA, and as being exposed to hepatitis C during follow-up if they had tested positive for HCV-RNA.\n\nWe used Cox regression models for time to event analyses, measuring time from enrollment to the diagnosis of a NADM, the last clinic visit in the study period, or death, whichever came first. Follow-up time in patients enrolled before 1996, the year cART was introduced, was left truncated at 31 December 1995. Patients diagnosed with a NADM before 1 January 1996 were excluded from the analysis for that specific NADM.\n\nIn a first step, we modeled the association of the time-updated immunological and virological exposure variables (CD4 cell count, CD8 cell count, CD4/CD8 ratio, and HIV viral load) with each of the four cancers separately, adjusting for the other variables. To reduce the risk of detecting associations that reflect reverse causality, we lagged these variables for 12 months or more, and captured cumulative exposure using simple moving averages (SMA). We entered the following representations of each variable: lagged by 12, 24 or 36 months; SMA over 12 months - lagged by 12 or 24 months; SMA over 24 months - lagged by 12 months (see Figure 1). In a next step we selected variables and representations of variables for inclusion into a final, malignancy-specific model. We selected one representation of each immunological and virological variable based on the goodness of fit (indexed by the Akaike information criteria [AIC]) and carried over variables significantly associated with cancer incidence (P <0.05) when modeled separately. The intermediary and final models were adjusted for the variables age, transmission group, smoking, education, chronic hepatitis B, exposure to hepatitis C, and calendar period. The time-updated variables for bacterial pneumonia (used only in the lung cancer analysis), hepatitis B, and hepatitis C were lagged by 12 months.\n\nWe assessed proportionality assumptions of the Cox model using the Schoenfeld residuals. Results are shown as hazard ratios (HR) with 95% confidence intervals (95% CI). We imputed missing values for smoking using multivariate imputation by chained equations and pooled results of five imputed dataset. We present the analyses with imputation since results were similar to complete-case analyses. All analyses were done in R Project for Statistical Computing (version 3.6.0) software24 using packages survival (version 2.42), mice (version 3.30) and mitools (version 2.3). R code and explanatory documents are available as extended data25.\n\n\nResults\n\nA total of 563 incident NADM were identified in the two data sources (SHCS and cancer registries). The SHCS recorded 60.9% of all cases, and the registries 76.2%. For the cancers analyzed in this study, the completeness of recording of cases in the SHCS was 54.3% for anal cancer, 66.7% for liver cancer, 71.4% for lung cancer, and 72.7% for prostate cancer. The corresponding percentages for the cancer registries were 85.7%, 80.6%, 85.7% and 93.2%, respectively (Table 1).\n\n* calculated as (c-a)/c\n\n& calculated as (c-b)/c\n\nAnal cancer was diagnosed in 70 patients over 71,592 person-years for an incidence rate of 9.78 per 10,000 person-years (95% confidence interval [CI] 7.62-12.35). Lung cancer was diagnosed in 49 patients over 71,888 person-years (6.82 per 10,000; 95% CI 5.04-9.01), prostate cancer in 44 men over 50,322 person-years (8.74 per 10,000; 95% CI 6.35-11.74), and liver cancer in 36 patients over 71,911 person-years (5.01 per 10,000; 95% CI 3.51-6.93).\n\nCompared to the Swiss general population, anal, lung, and liver cancer all exhibited higher rates in the SHCS, whereas the rate of prostate cancer was similar (Figure 2). The difference was largest for anal cancer: the incidence was 76 times higher in people with HIV (SIR 76.1, 95% CI 60.2-96.2) than in the general Swiss population.\n\nCharacteristics of SHCS participants who developed cancer and participants who did not are shown in Table 2. Compared to those who remained free of cancer, HIV-positive individuals, who developed cancer were slightly older, more likely to be smokers, and at a more advanced clinical stage of HIV infection according to the Centers for Disease Control and Prevention classification26. The distribution across HIV transmission groups also varied between cancers: the majority of patients who developed anal or prostate cancer were MSM, while those with lung or liver cancer were predominantly patients with a history of IDU.\n\nSD, standard deviation; cART, combination anti-retroviral therapy; CDC, Centers for Disease Control and Prevention; MSM, men having sex with men; IDU, intravenous drug use.\n\nWe selected one representation of each exposure variable based on goodness of fit models. The AIC values for the different models are shown in Table 3, the results of the separate models for each immunological and virological variable in Table 4 and the results of the mutually adjusted models in Table 5. For anal cancer, CD4 count, CD8 count and CD4/CD8 ratio were all associated with cancer risk in the separate models (Table 4). In the mutually adjusted model, a higher CD4 cell count (lagged 12 months) and a CD8 count below 1000 cells/µl (lagged by 24 months) continued to be significantly associated (P=0.002) with a reduced risk of anal cancer (Table 5). For example, in patients with a 12-month lagged CD4 count above 500 cells/µl, risk was reduced to about a fourth compared to individuals with a CD4 count below 200 cells/µl (hazard ratio [HR] 0.24; 95% CI 0.10-0.61). The risk of anal cancer was substantially higher in MSM (HR 6.86, 95% CI 1.64-28.6) than other men and higher in smokers (HR 2.07, 95% CI 1.07-3.98) compared to non-smokers.\n\nLag12, lagged by 12 months; SMA12: simple moving average over 12 months.\n\nModels are adjusted for age, HIV transmission group, smoking, hepatitis B and C statuses, education, and calendar year, and include only one exposure variable. Numbers in bold highlight the representations selected for analysis.\n\na estimate for CD4 cell count 200-499 cells/µl. b the reference category is <500 HIV RNA copies per ml.\n\nHR, hazard ratio; CI, confidence interval; SMA, simple moving average; na, not applicable. Models for anal and liver cancer are adjusted for age, HIV transmission group, smoking, hepatitis B and C status, education, and calendar period. Models for lung cancer are stratified by age and adjusted for HIV transmission group, smoking, hepatitis C status, education, and calendar period. Models for prostate cancer are stratified for age and adjusted for HIV transmission group, smoking, hepatitis B and C status, education, and calendar period.\n\na estimate for CD4 cell count 200-499, b estimate for other women or men\n\nHR, hazard ratio; SMA, simple moving average; MSM, men having sex with men.\n\nAll models are adjusted for the variables shown and for age, calendar period, and highest completed education. Models for lung cancer and prostate cancer are stratified with respect to age.\n\nFor lung cancer and prostate cancer, Cox regression models were stratified by age to preserve proportional hazards. In the case of lung cancer, CD8 cell count (lagged by 36 months) was the only significant predictor (P=0.04) among the immunological and virological variables (Table 2): participants with CD8 cell counts equal to or above 1000 cells/µl had a greater risk of developing lung cancer, compared to those with counts below 1000 cells/µl (HR 1.89, 95% CI 1.03-3.48). A history of bacterial pneumonia was associated with lung cancer in the separate model (Table 4), but not after adjusting for CD8 cell count (Table 5). Smoking at enrollment was strongly associated with lung cancer (HR 6.50, 95% CI 1.36-31.2, Table 5). For prostate cancer, none of the immunological or virological were associated with a risk of cancer (Table 4), though there was some evidence of smoking being associated with higher rates (HR 1.92, 95% CI 0.96-3.83, Table 5).\n\nFor liver cancer, exposure to lower CD4 counts (SMA over 24 months, lagged by 12 months) and the CD4/CD8 ratio (lagged by 24 months) were associated with cancer incidence in the separate models (Table 4). In the final model (Table 5), only the CD4 cell count was predictive, with patients with average counts of 500 cells/µl or above at much lower risk than patients with counts below 200 cells/µL (HR 0.07, 95% CI 0.01-0.59). Chronic hepatitis B infection (HR 7.86, 95% CI 3.78-16.4) and hepatitis C infection (HR 4.77, 95% CI 1.80-12.7) were both strongly associated with liver cancer (Table 5).\n\n\nDiscussion\n\nIn the era of cART, life expectancy of PLWHIV is approaching that of the general population, the incidence of AIDS-defining malignancies has declined sharply, and NADM have emerged as an important co-morbidity1,2,27,28. This study linked the database of the SHCS with three cantonal cancer registries to identify incident NADM. Anal, lung, prostate and liver cancers were common NADM among participants of SHCS. Compared to the general population, three of the four cancers occurred more frequently among PLWHIV. The exception was prostate cancer, whose incidence was similar to the general population. Associations of lagged and cumulative exposure to immunodeficiency or immune senescence were evident for some but not all cancers, and HIV replication was not related to cancer risk.\n\nThe SHCS is a well-established cohort of people living with HIV in Switzerland, where regular follow-up visits provide a wealth of information to inform fundamental, clinical and epidemiological research17,18. Both genders and the various transmission groups are well represented. We used state-of-the-art privacy preserving linkage to complement incident cancers recorded in the SHCS with cases from cantonal registries19,20. Interestingly, neither the SHCS nor the cancer registries had a complete record of cases. This situation should improve in the near future, with the introduction of nationwide, compulsory cancer registration in Switzerland in 202029. Previous analyses of cancer risk in the SHCS used a matched case-control approach, which precluded the study of some risk factors30,31. For example, matching cancer cases and controls on transmission group hampered the evaluation of hepatitis virus infections for liver cancer31, or of smoking for anal cancer32. Our study also has limitations. Cancer linkage was only possible for three cantonal registries, which meant that about 35% of SHCS participants could not be included, limiting statistical power. Anal intercourse is a risk factor for anal cancer, and excessive alcohol consumption for liver cancer, but data on sexual behavior and alcohol have been collected in the SHCS only since 2000 and 2005, respectively, and were not considered here.\n\nThe reasons for the higher cancer incidence in PLWHIV differ across cancers, and include multiple factors. Several studies have shown that immunodeficiency was associated with a higher risk of NADM28,33–35. Some of these studies analyzed NADM as a single group, or as groups of infection-related and non-infection-related NADM. We found that immunodeficiency (indexed by exposure to low CD4 counts) was associated with anal cancer and liver cancer, but not with lung cancer or prostate cancer. Our results indicate that each NADM should be analyzed separately, rather than combining different NADM into groups.\n\nFor lung cancer, a high CD8 cell count several years prior to the diagnosis was a risk factor whereas a low CD4 count was not associated with lung cancer. Persistent CD8 cell elevation among HIV patients on long-term cART, a marker of immune senescence, is associated with inflammatory, non-AIDS related events such as cardiovascular, renal, respiratory, metabolic diseases and NADM, and with increased non-AIDS-related mortality36–38. Mussini and colleagues suggested that the CD4/CD8 ratio might be a more robust marker of the immune dysfunction associated with NADM than the CD4 or the CD8 cell count39. Indeed, a study of US veterans living with HIV found that the CD4/CD8 ratio was the strongest predictor of incidence with CD4/CD8 ratio ratios below 1.0 associated with higher risk15. In contrast to the veterans study, the CD8 cell count was more robustly associated with lung cancer risk than either the CD4 cell count or the CD4/CD8 ratio in our analysis, suggesting that in the cART era, immune senescence rather than immunodeficiency increases the risk of lung cancer.\n\nA history of recurrent pneumonia was associated with lung cancer risk in the HIV/AIDS Cancer Match study40 and in the US Veterans Aging Cohort Study15. In our analyses, the time-updated history of bacterial pneumonia was associated with lung cancer risk in the model adjusted for baseline variables, but evidence of this association disappeared after adjusting for CD8 count. Clearly, the most important risk factor for lung cancer is smoking, which in Switzerland and elsewhere is substantially more common among PLWHIV than in the general population41–43. Smoking was also associated with anal cancer, but it remains unclear whether the association is causal or whether smoking is a marker for unsafe sexual behavior and exposure to HPV. Several studies have found that smoking may be associated with a higher risk of HPV infection44–46.\n\nProstate cancer is one of the most common malignancies among men in Western populations, and an important cause of cancer death. Older age, a positive family history, black race and elevated levels of insulin growth factor (IGF)-I have been shown to be risk factors for prostate cancer47,48. Our results do not suggest that immunodeficiency or immune senescence strongly influence prostate cancer risk. In this context, it is noteworthy that prostate cancer is not consistently responsive to immune therapy, in contrast to other solid tumors, including lung cancer49.\n\nLiver cancer was associated with prolonged exposure to CD4 counts below 500 cells/µl however; chronic viral hepatitis was a likely cause in almost all patients developing liver cancer: there was evidence of chronic HBV or HCV infection in 83% of patients with liver cancer. HBV and HCV were both strong risk factors for the cancer. An explanation for the strong association with HBV might be hepatitis delta virus (HDV) co-infection, which in the SHCS is observed in about 15% of patients with chronic HBV infection50. HDV infection accelerates the progression of HBV related liver disease, including hepatocellular cancer50.\n\nImmunodeficiency and, to a lesser extent, immune senescence are prevented by starting cART as soon as possible after HIV diagnosis36, as recommended by the World Health Organization51. The “Universal Test and Treat” approach to HIV coupled with interventions to ensure adherence will likely reduce the incidence of some NADM. The integration of smoking cessation interventions into routine care of people living with HIV is another important measure52. All SHCS participants are tested for chronic HBV infection to guide decisions on HBV vaccination and on the inclusion antiretrovirals with activity against HBV in cART regimens, as recommended by guidelines53,54. Similarly, testing for antibodies against HCV is done at HIV diagnosis and annually thereafter, and there is universal access to direct acting anti-virus (DAA) therapy against HCV. Future analyses of the incidence of NADM in the SHCS will document the impact of these measures. Of note, the effectiveness of anal cancer screening is unclear: the ongoing Anal Cancer/HSIL Outcomes Research Study (ANCHOR) aims to determine whether treatment of high grade squamous intraepithelial lesions (HSIL) prevents anal cancer55.\n\n\nConclusions\n\nThe importance of immunodeficiency and immune senescence, and of other risk factors differs across NADM, with important implications for prevention. Immunodeficiency was an important risk factor for anal and liver cancer whereas immune senescence was associated with lung cancer and anal cancer.\n\nThe SHCS was approved by the Ethics committees of the participating institutions. At enrolment, SHCS participants provide written informed consent for the use of biological and clinical data. For the present study, we obtained additional ethical approval for privacy-preserving probabilistic record linkage from the Ethics Committee of the Canton of Bern.\n\n\nData availability\n\nThe National Institute for Cancer Epidemiology and Registration (NICER) is happy to share the anonymized cancer registry data with eligible partners. If you wish to receive specific data from the NICER database, please complete the respective form at https://www.nicer.org/en/data/request-data/.\n\nThe data analyzed in this study are sensitive, having been contributed by people living with HIV. Access to de-identified data is restricted to collaborative projects that have been submitted to and approved by the Scientific Board of the SHCS. Please contact the SHCS at http://www.shcs.ch/contact if you are interested in gaining access to the study data. Sharing or linkage of data may require additional ethics approval.\n\nFiles with the R code used in the analysis are provided as extended data, along with an explanatory document, on Open Science Framework\n\nOpen Science Framework: Extended data for Non-AIDS defining malignancies in the combination ART era: immunological and socio-behavioral risk factors (Ruffieux et al. Faculty1000research 2019). https://doi.org/10.17605/OSF.IO/GY5VM25\n\nThis project contains the following extended data:\n\ncode.description final.pdf (description of R code)\n\npilot.R (initializes data preprocessing for a given cancer)\n\npreprocess_SHCS.R (loads SHCS data)\n\npreprocess_linkage.R (loads the cases of cancer that could be linked to the SHCS)\n\nmerge.R (appends patient-level information, creates dataset with one line per patient)\n\nrisk_factors_monthly.R (creates dataset with one line per month, per patient, with time-updated variables)\n\nwrite_res_df.R (sets up data for survival analysis for a specific cancer)\n\ncancer_counts.R (produces the number of cancers per data source reported in Table 1)\n\npatient_characteristics.R (produces results in Table 2)\n\ncrude_incidence_rates.R (produces crude cancer incidence rates)\n\nSIR_NADCs.R (produces the SMRs shown in Figure 2)\n\nimpute_smoking.R (multiple imputation of smoking variable)\n\nCox_model_selection.R (produces the AICs in Table 3).\n\nrisk_factors_Cox.R (Cox regression to produce results in Table 4 and Table 5)\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nThe linkage and analysis of the data were funded by grants from Cancer Research Switzerland [KFS-3165-02-2013 and KFS-3862-02-2016]. The SHCS is supported by the Swiss National Science Foundation (SNSF) [177499]. ME is supported by special project funding from the SNSF [174281].\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe are grateful to the patients, study nurses, and care providers who participate in the SHCS.\n\nThe members of the Swiss HIV Cohort Study are: Aebi-Popp K, Anagnostopoulos A, Battegay M, Bernasconi E, Böni J, Braun DL, Bucher HC, Calmy A, Cavassini M, Ciuffi A, Dollenmaier G, Egger M, Elzi L, Fehr J, Fellay J, Furrer H, Fux CA, Günthard HF (President of the SHCS), Haerry D (deputy of \"Positive Council\"), Hasse B, Hirsch HH, Hoffmann M, Hösli I, Huber M, Kahlert CR (Chairman of the Mother & Child Substudy), Kaiser L, Keiser O, Klimkait T, Kouyos RD, Kovari H, Ledergerber B, Martinetti G, Martinez de Tejada B, Marzolini C, Metzner KJ, Müller N, Nicca D, Paioni P, Pantaleo G, Perreau M, Rauch A (Chairman of the Scientific Board), Rudin C, Scherrer AU (Head of Data Centre), Schmid P, Speck R, Stöckle M (Chairman of the Clinical and Laboratory Committee), Tarr P, Trkola A, Vernazza P, Wandeler G, Weber R, Yerly S.\n\n\nReferences\n\nWandeler G, Johnson LF, Egger M: Trends in life expectancy of HIV-positive adults on antiretroviral therapy across the globe: comparisons with general population. Curr Opin HIV AIDS. 2016; 11(5): 492–500. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGueler A, Moser A, Calmy A, et al.: Life expectancy in HIV-positive persons in Switzerland: matched comparison with general population. AIDS. 2017; 31(3): 427–36. PubMed Abstract \| Free Full Text\n\nPatterson S, Cescon A, Samji H, et al.: Life expectancy of HIV-positive individuals on combination antiretroviral therapy in Canada. BMC Infect Dis. 2015; 15: 274. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSmith CJ, Ryom L, Weber R, et al.: Trends in underlying causes of death in people with HIV from 1999 to 2011 (D:A:D): a multicohort collaboration. Lancet. 2014; 384(9939): 241–8. PubMed Abstract \| Publisher Full Text\n\nAntiretroviral Therapy Cohort Collaboration: Causes of death in HIV-1-infected patients treated with antiretroviral therapy, 1996-2006: collaborative analysis of 13 HIV cohort studies. Clin Infect Dis. 2010; 50(10): 1387–96. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMorlat P, Roussillon C, Henard S, et al.: Causes of death among HIV-infected patients in France in 2010 (national survey): trends since 2000. AIDS. 2014; 28(8): 1181–91. PubMed Abstract \| Publisher Full Text\n\nSchwarcz SK, Vu A, Hsu LC, et al.: Changes in causes of death among persons with AIDS: San Francisco, California, 1996-2011. AIDS Patient Care STDS. 2014; 28(10): 517–23. PubMed Abstract \| Publisher Full Text\n\nFarahani M, Mulinder H, Farahani A, et al.: Prevalence and distribution of non-AIDS causes of death among HIV-infected individuals receiving antiretroviral therapy: a systematic review and meta-analysis. Int J STD AIDS. 2017; 28(7): 636–50. PubMed Abstract \| Publisher Full Text\n\nGrulich AE, van Leeuwen MT, Falster MO, et al.: Incidence of cancers in people with HIV/AIDS compared with immunosuppressed transplant recipients: a meta-analysis. Lancet. 2007; 370(9581): 59–67. PubMed Abstract \| Publisher Full Text\n\nEngels EA: Non-AIDS-defining malignancies in HIV-infected persons: etiologic puzzles, epidemiologic perils, prevention opportunities. AIDS. 2009; 23(8): 875–85. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBorges AH, Dubrow R, Silverberg MJ: Factors contributing to risk for cancer among HIV-infected individuals, and evidence that earlier combination antiretroviral therapy will alter this risk. Curr Opin HIV AIDS. 2014; 9(1): 34–40. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nClifford GM, Polesel J, Rickenbach M, et al.: Cancer risk in the Swiss HIV Cohort Study: associations with immunodeficiency, smoking, and highly active antiretroviral therapy. J Natl Cancer Inst. 2005; 97(6): 425–32. PubMed Abstract \| Publisher Full Text\n\nBorges ÁH, Silverberg MJ, Wentworth D, et al.: Predicting risk of cancer during HIV infection: the role of inflammatory and coagulation biomarkers. AIDS. 2013; 27(9): 1433–41. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDeeks SG: HIV infection, inflammation, immunosenescence, and aging. Annu Rev Med. 2011; 62(1): 141–55. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSigel K, Wisnivesky J, Crothers K, et al.: Immunological and infectious risk factors for lung cancer in US veterans with HIV: a longitudinal cohort study. Lancet HIV. 2017; 4(2): e67–73. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHema MN, Ferry T, Dupon M, et al.: Low CD4/CD8 Ratio Is Associated with Non AIDS-Defining Cancers in Patients on Antiretroviral Therapy: ANRS CO8 (Aproco/Copilote) Prospective Cohort Study. Apetrei C, editor. PLoS One. 2016; 11(8): e0161594. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLedergerber B, von Overbeck J, Egger M, et al.: The Swiss HIV Cohort Study: rationale, organization and selected baseline characteristics. Soz Praventivmed. 1994; 39(6): 387–94. PubMed Abstract \| Publisher Full Text\n\nSwiss HIV Cohort Study, Schoeni-Affolter F, Ledergerber B, et al.: Cohort profile: the Swiss HIV Cohort study. Int J Epidemiol. 2010; 39(5): 1179–89. PubMed Abstract \| Publisher Full Text\n\nSchmidlin K, Clough-Gorr KM, Spoerri A, et al.: Privacy preserving probabilistic record linkage (P3RL): a novel method for linking existing health-related data and maintaining participant confidentiality. BMC Med Res Methodol. 2015; 15: 46. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchnell R, Bachteler T, Reiher J: Privacy-preserving record linkage using Bloom filters. BMC Med Inform Decis Mak. 2009; 9: 41. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchmidlin K, Clough-Gorr KM, Spoerri A, et al.: Privacy preserving probabilistic record linkage (P3RL): a novel method for linking existing health-related data and maintaining participant confidentiality. BMC Med Res Methodol. 2015; 15: 46. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNational Institute for Cancer Epidemiology and Registration [Internet]. Reference Source\n\nBreslow N, Day N: Statistical methods in cancer research. Volume II: The design and analysis of cohort studies. Vol. 2, IARC Scientific Publications No. 82. Lyon, France: Oxford University Press; 1987. Reference Source\n\nR Development Core Team: R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 2010. Reference Source\n\nEgger M: Extended data for Non-AIDS defining malignancies in the combination ART era: immunological and socio-behavioral risk factors (Ruffieux et al. Faculty1000research 2019). 2019. http://www.doi.org/10.17605/OSF.IO/GY5VM\n\nAcquired immunodeficiency syndrome (AIDS). 1987 revision of CDC/WHO case definition for AIDS. Wkly Epidemiol Rec. 1988; 63: 1–7. Reference Source\n\nShiels MS, Engels EA: Evolving epidemiology of HIV-associated malignancies. Curr Opin HIV AIDS. NIH Public Access; 2017; 12: 6–11. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMazzotta E, Tontodonati M, Gabrielli C, et al.: Prevalence and predictors of malignancies in a polycentric cohort of HIV patients from Italy. J Int AIDS Soc. 2014; 17(4 Suppl 3): 19652. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOffice fédéral de la santé publique OFSP: Enregistrement du cancer [Internet]. [cited 2019 Jan 6]. Reference Source\n\nClifford GM, Lise M, Franceschi S, et al.: Lung cancer in the Swiss HIV Cohort Study: role of smoking, immunodeficiency and pulmonary infection. Br J Cancer. 2012; 106(3): 447–52. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nClifford GM, Rickenbach M, Polesel J, et al.: Influence of HIV-related immunodeficiency on the risk of hepatocellular carcinoma. AIDS. 2008; 22(16): 2135–41. PubMed Abstract \| Publisher Full Text\n\nBertisch B, Franceschi S, Lise M, et al.: Risk factors for anal cancer in persons infected with HIV: a nested case-control study in the Swiss HIV Cohort Study. Am J Epidemiol. 2013; 178(6): 877–84. PubMed Abstract \| Publisher Full Text\n\nPacheco YM, Jarrin I, Rosado I, et al.: Increased risk of non-AIDS-related events in HIV subjects with persistent low CD4 counts despite cART in the CoRIS cohort. Antiviral Res. 2015; 117: 69–74. PubMed Abstract \| Publisher Full Text\n\nKesselring A, Gras L, Smit C, et al.: Immunodeficiency as a risk factor for non-AIDS-defining malignancies in HIV-1-infected patients receiving combination antiretroviral therapy. Clin Infect Dis. 2011; 52(12): 1458–65. PubMed Abstract \| Publisher Full Text\n\nGuiguet M, Boué F, Cadranel J, et al.: Effect of immunodeficiency, HIV viral load, and antiretroviral therapy on the risk of individual malignancies (FHDH-ANRS CO4): a prospective cohort study. Lancet Oncol. 2009; 10(12): 1152–9. PubMed Abstract \| Publisher Full Text\n\nCao W, Mehraj V, Kaufmann DE, et al.: Elevation and persistence of CD8 T-cells in HIV infection: the Achilles heel in the ART era. J Int AIDS Soc. 2016; 19(1): 20697. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTrickey A, May MT, Schommers P, et al.: CD4:CD8 Ratio and CD8 Count as Prognostic Markers for Mortality in Human Immunodeficiency Virus-Infected Patients on Antiretroviral Therapy: The Antiretroviral Therapy Cohort Collaboration (ART-CC). Clin Infect Dis. 2017; 65(6): 959–66. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHelleberg M, Kronborg G, Ullum H, et al.: Course and Clinical Significance of CD8+ T-Cell Counts in a Large Cohort of HIV-Infected Individuals. J Infect Dis. 2015; 211(11): 1726–34. PubMed Abstract \| Publisher Full Text\n\nMussini C, Lorenzini P, Cozzi-Lepri A, et al.: CD4/CD8 ratio normalisation and non-AIDS-related events in individuals with HIV who achieve viral load suppression with antiretroviral therapy: an observational cohort study. Lancet HIV. 2015; 2(3): e98–106. PubMed Abstract \| Publisher Full Text\n\nShebl FM, Engels EA, Goedert JJ, et al.: Pulmonary infections and risk of lung cancer among persons with AIDS. J Acquir Immune Defic Syndr. 2010; 55(3): 375–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCalvo-Sánchez M, Perelló R, Pérez I, et al.: Differences between HIV-infected and uninfected adults in the contributions of smoking, diabetes and hypertension to acute coronary syndrome: two parallel case-control studies. HIV Med. 2013; 14(1): 40–8. PubMed Abstract \| Publisher Full Text\n\nCrothers K, Goulet JL, Rodriguez-Barradas MC, et al.: Impact of cigarette smoking on mortality in HIV-positive and HIV-negative veterans. AIDS Educ Prev. 2009; 21(3 Suppl): 40–53. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGmel G, Kuendig H, Maffli E, et al.: Suchtmonitoring Schweiz. Jahresbericht - Daten 2011. Bern; 2012. Reference Source\n\nVaccarella S, Herrero R, Snijders PJ, et al.: Smoking and human papillomavirus infection: pooled analysis of the International Agency for Research on Cancer HPV Prevalence Surveys. Int J Epidemiol. 2008; 37(3): 536–46. PubMed Abstract \| Publisher Full Text\n\nWieland U, Hellmich M, Wetendorf J, et al.: Smoking and anal high-risk human papillomavirus DNA loads in HIV-positive men who have sex with men. Int J Med Microbiol. 2015; 305(7): 689–96. PubMed Abstract \| Publisher Full Text\n\nCombes JD, Heard I, Poizot-Martin I, et al.: Prevalence and Risk Factors for Anal Human Papillomavirus Infection in Human Immunodeficiency Virus-Positive Men Who Have Sex with Men. J Infect Dis. 2018; 217(10): 1535–43. PubMed Abstract \| Publisher Full Text\n\nGrönberg H: Prostate cancer epidemiology. Lancet. 2003; 361(9360): 859–64. PubMed Abstract \| Publisher Full Text\n\nRenehan AG, Zwahlen M, Minder C, et al.: Insulin-like growth factor (IGF)-I, IGF binding protein-3, and cancer risk: systematic review and meta-regression analysis. Lancet. 2004; 363(9418): 1346–53. PubMed Abstract \| Publisher Full Text\n\nSlovin SF: Immunotherapy for prostate cancer: is prostate an immune responsive tumor? Curr Opin Urol. 2016; 26(6): 529–34. PubMed Abstract \| Publisher Full Text\n\nBéguelin C, Moradpour D, Sahli R, et al.: Hepatitis delta-associated mortality in HIV/HBV-coinfected patients. J Hepatol. 2017; 66(2): 297–303. PubMed Abstract \| Publisher Full Text\n\nWorld Health Organization: Consolidated Guidelines on the Use of Antiretroviral Drugs for Treating and Preventing HIV Infection: Recommendations for a Public Health Approach. Second Edition 2016. Geneva; 2016. Reference Source\n\nGoncalves PH, Montezuma-Rusca JM, Yarchoan R, et al.: Cancer prevention in HIV-infected populations. Semin Oncol. 2016; 43(1): 173–88. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMasur H, Kaplan JE, Holmes KK, et al.: Guidelines for preventing opportunistic infections among HIV-infected persons--2002. Recommendations of the U.S. Public Health Service and the Infectious Diseases Society of America. Ann Intern Med. 2002; 137(5 Pt 2): 435–78. PubMed Abstract \| Publisher Full Text\n\nEACS European AIDS Clinical Society: Guidelines. Version 9.1. October 2018. 2018. Reference Source\n\nAnchor Trial Launch. National Cancer Institute [Internet]. [cited 2019 Jan 7]. Reference Source" }	[ { "id": "52261", "date": "03 Sep 2019", "name": "Jean-Pierre Routy", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nNon-AIDs cancers seem to be more frequent in well-treated patients on ART than the general population, due either to HIV persistent immune damage vs. lifestyle, behaviour risk or both. To prove this link, HIV infected persons should be compared to age matched general population in a relatively large population as the events remain encouragingly relatively low, during the ART era, where viral replication is controlled over long periods of time. To this end, investigators from the well-established HIV Swiss cohort calculated standardized incidence ratios for each identified NADM by comparing the observed number of cancers with the expected number of occurrences. An indirect standardization was performed using period-specific rates for the Swiss population provided by the National Institute for Cancer Epidemiology and Registration using age and sex as co-variables. Influence of CD4 and CD8 count, CD4/CD8 ratio, and HIV viral load was assessed by Cox regression models. Higher rates of HPV driven anal (SIR 76.1, 95% Confidence interval (CI) 60.2-96.2), lung (SIR 1.98, 1.50-2.62), and liver cancer (SIR 7.28, 5.25-10.1) were observed.\n\nHormonally driven cancer like prostate cancer was similar in HIV and control population (SIR 1.03, 0.76-1.38). Interestingly, anal cancer was associated with low CD4, with high CD8, and smoking. Interestingly, for lung cancer, CD8 cell count was the only predictor identified opening new hypothesis on inflammation in the context of ART-treated persons. As expected, CD4 cell count and chronic hepatitis B and C infection were predictive of liver cancer incidence.\n\nSeveral issues should be addressed:\nTypes of cancer should be divided into viral-related and those not known to be associated with viruses like lung and prostate.\n\nFocus should be on lung cancer and its relationship with CD8 and nadir CD4 T cell if data available.\n\nEffort should be on novelty and not confirming previous data.\n\nChanges occurring or not in cancer incidence during the study period should be highlighted.\n\nDiscussion on Cd8 elevation and occurrence of lung cancer should be more detailed as future direction for research.\n\nConclusion:\nImmunodeficiency was an important risk factor for anal and liver cancer, whereas immune senescence was associated with lung cancer and anal cancer. I think anal cancer with immune senescence should be removed from the conclusion as it seems to be more driven by CD4 T cells and the conclusion is not clear. Findings of lung cancer and immune senescence are the novelty of this manuscript and open several hypothesis including smoking and CD8 elevation in non-HIV population and immune senescence linked with CMV serostatus.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "52874", "date": "23 Sep 2019", "name": "Keith Sigel", "expertise": [ "Reviewer Expertise HIV and cancer risk/outcomes" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting and well written manuscript describing a study evaluating the incidence of non-AIDS cancers as well as association of longitudinal measures of immunologic and infectious exposures in the Swiss HIV Cohort. The study identifies immunologic risk factors for several NADCs, some of which differ from previous studies. The authors found higher risks of several NADCs in HIV+ persons but not prostate cancer, generally similar findings to other large population-based studies. Specific immunologic disturbances were different for the more prevalent cancers (i.e., low CD4/high CD8 for anus). This analysis utilizes rigorous and well-described methods and is a useful comparison for other similar studies from other cohorts.\n\nIntroduction:\n\n\"This type of cancer is now the most .. \" -- this should probably be plural (these types of cancer)\n\nYou might want to define \"cantonal\" - it is not a well-known word\n\nDiscussion:\n\nYou say that \" immune senescence rather than immunodeficiency increases the risk of lung cancer.\" This needs additional justification and/or support (and likely is too strong of a statement) as you would need additional phenotyping of these CD8s (i.e., additional flow markers) to definitively say that there is a increase in senescence. This also may need to be tempered in the conclusion. If you can provide good evidence that excess CD8s indicates senescence (as opposed to activation or other phenotypic states) then that would also likely suffice.\n\nThis study had far fewer lung cancers than the VACS study - that should be mentioned; for the lung finding at the very least and potentially for any other findings that contract other large studies a power calculation should be performed to determine the minimum effect that could be detected for CD4, CD4/CD8 and/or pneumonia based on the numbers of cancer cases. Also for the lung findings, it would be useful to compare to the French studies that have evaluated similar immunologic exposures and found associations with CD4/CD8 and longitudinal CD4 level.\n\nLack of pack-year smoking exposure information should be included as a limitation.\n\nA minor criticism is that the data is not available for confirmation due to its sensitive nature. This is understandable and does not need to be addressed.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-1400
https://f1000research.com/articles/8-1399/v1	09 Aug 19	{ "type": "Case Report", "title": "Case Report: A rare case of BCGitis in a patient with bladder cancer treated with the BCG vaccine", "authors": [ "Domingos Sousa", "Sérgio Antunes Silva", "Catarina Jorge", "Ana Isabel Rodrigues", "Rita Martins Fernandes", "Margarida Viana Coelho", "Elena Rios", "Rita Ferro Rodrigues", "Amparo Mingo", "Sérgio Antunes Silva", "Catarina Jorge", "Ana Isabel Rodrigues", "Rita Martins Fernandes", "Margarida Viana Coelho", "Elena Rios", "Rita Ferro Rodrigues", "Amparo Mingo" ], "abstract": "The Bacillus Calmette Guérin (BCG) vaccine was initially produced as a vaccine against tuberculosis. BCG is an attenuated live strain of Mycobacterium bovis and has been widely used as an immunotherapy over the last years in bladder cancer. We describe a case of a 61-year-old Caucasian male with previous bladder cancer, who had been treated for the last 15 months with instillation of BCG, admitted with 3-week evolution history of fever (38ºC), asthenia, anorexia and a weight loss of 6kg. The patient’s condition deteriorated leading to sepsis. A liver biopsy was performed showing granulomatous dispersed through all the parenchyma. Empirical therapy for M. bovis was started with good response. Even though it is rare, BCGitis must be ruled out in any patient submitted to immunotherapy with BCG and prompt therapy started if disseminated disease is present, which improves the outcome.", "keywords": [ "BCGitis", "Mycobacterium bovis", "Bladder cancer", "sepsis" ], "content": "Introduction\n\nMost bladder carcinoma are located superficially, described as a non-muscle invasive bladder cancer, which allows a therapeutic approach with local immunotherapy1. The Bacillus Calmette Guérin (BCG) vaccine was initially produced as a vaccine against tuberculosis. BCG is an attenuated live strain of Mycobacterium bovis and has been widely used as an immunotherapy over the last few years2. Treatment with the BCG vaccine is well tolerated by over 95% of patients3.\n\nThe most frequent side effects of immunotherapy with the BCG vaccine are limited to the bladder – bacterial or chemical cystitis and haematuria. Although rare, systemic side effects can be mild or severe and they range from fever and influenza-like symptoms to BCG-induced lung infection, liver toxicity and BCG sepsis. Sepsis was estimated to affect 0.3% of patients treated with BCG instillation3. When a disseminated infection is present it can be described as BCGitis, with the lung and liver being the most frequently affected organs (50% and 61%, respectively)4\n\nA diagnosis of BCGitis can be validated by direct staining (Ziehl-Neelsen or auramine-rhodamine), histological observation (presence of a granulomatous pattern with or without caseous necrosis) or the results of polymerase-chain reaction (PCR) assays. However, the rentability of all options is low, leading to increasing difficultly in establishing the diagnosis4,5.\n\nWe hereby report a rare case of septic BCGitis.\n\n\nCase report\n\nA 61-year-old Caucasian male with a medical history of pulmonary embolism and bladder cancer (pTaG1), who had been treated with the BCG vaccine for the last 15-months (a total of 15 instillations) was admitted with a 3-week history of fever (38°C), asthenia, anorexia and a weight loss of 6kg.\n\nA week prior to admission (virgula) the patient presented with dysuria and pollakiuria, justifying an empirical treatment for a urinary tract infection with ciprofloxacin (500mg bid) but without response. the urine culture requested came back negative.\n\nThe patient’s general state was aggravated with persistent fever, tachycardia and increased inflammatory markers (CRP 46mg; normal range, <5mg/L), elevated liver enzymes [AST (46U/l); ALT (64U/I)], leading to hospital admission with a diagnosis of sepsis (Day 1; Table 1). An empirical broad-spectrum course of antibiotics with ceftriaxone (2gr) and gentamycin (500mg) was instituted for 5 days. A septic screen was performed, which grew no microorganisms.\n\nHaemoglobin (Hb) is expressed in grams per liter; Neutrophil cells, platelets, lymphocytes and monocytes are expressed as elements /mm³; Aspartate transaminase (AST), alanine transaminase (ALT), gamma-glutamyl transferase (gGT), and alkaline phosphatase (ALP) are all expressed in international units per liter; Total bilirubin is expressed in milligrams per decilitre and C-reactive protein is expressed in milligrams per litre.\n\nAfter 3 days of hospital stay, the patient’s status continued to deteriorate with persistent fever, hypotension and an increase in CRP to 76mg/L (Day 3; Table 1). No abnormal lymph nodes were observed. An abdominal ultrasound revealed a hyperechoic liver lesion with preservation of length and liver contour (Figure 1). This evidence (virgula) combined with the knowledge that there had been a previous recent administration of immunotherapy with BCG 3 weeks before the onset of the presented clinical symptoms, lead physicians to perform a liver biopsy.\n\nA liver biopsy was conducted revealing multiple small epithelial granulomas without caseous necrosis compatible with granulomatous hepatitis. Ziehl-Neelsen staining and Löwenstein–Jensen medium culture were negative. A PCR assay for M. tuberculosis complex (MTB) was also negative both in the liver biopsy and urine sample.\n\nDue to continuous clinical worsening of the patient’s condition until day 7 and aggravated sepsis with persistent fever, hypotension and increased CRP to 186mg/dL under broad-spectrum antibiotics, an empirical regimen with isoniazid 300mg, rifabutin 300mg and ethambutol 1200mg was started for 3 months. Given its interaction with rivaroxaban, rifampicin was switched to rifabutin. In a matter of days, the clinical outcome with resolution of the fever and normalized the inflammation markers at day seven from admission (Table 1). Moreover, within one week of therapy, CRP, AST and ALT revealed a notorious decrease, with a complete normalization within 11 days upon treatment institution.\n\nThe patient´s therapeutic regimen included isoniazid 300mg, rifabutin 300mg and ethambutol 1200mg for the first 3 months and the subsequent 3 months with isoniazid and rifabutin with good adherence and physiological tolerance. After 2 years of treatment, the patient remains in remission.\n\n\nDiscussion\n\nOur case highlights the rare but dangerous systemic secondary side effects of BCG instillations that characterize (singular) BCGitis. Evolution to a multisystemic disease deserves attention because its diagnosis can be a challenge to any physician – unspecified symptoms could also be present in acute infection, autoimmune disorders, evolution of malignancy or drug toxicity.\n\nThe pathophysiology of BCGitis is not yet well understood. Previous studies have suggested that it is the result from hypersensitivity, with a diffuse granulomatous reaction without isolation of M. bovis, while others consider it a disseminated disease of M. bovis4.\n\nThe identification of Mycobacterium requires a combination of tests because there is a poor reliability of PCR for MTB, Ziehl-Neelsen or auramine-rhodamine stains and also with Löwenstein–Jensen medium culture4,6. The histological finding of granulomatous lesions remains the main support to the diagnosis when it is not possible to identify Mycobacteria, as we describe in this case. However, the steatotic liver pattern presented in Figure 1 supports the presence of liver inflammation and the need of a liver biopsy.\n\nIn sensitive strains, the use of empirical antibiotherapy with fluoroquinolone could decrease the accuracy of Ziehl-Neelsen stain, Löwenstein–Jensen medium culture and PCR for MTB.\n\nOptions for treatment regimens and duration of therapy also remain under discussion; however, treatment options that have been accepted vary from 6–9 months4,5. Our patient was treated for 6 months, which proved to be adequate for treatment of BCGitis. The interaction of rivaroxaban with rifampicin forced a switch to rifabutin.\n\nCorticosteroids should only be used in severe cases, due to the lack of evidence in mild and moderate presentations4.\n\nFinally, the present case confirms the good clinical and laboratorial response upon institution of empirical therapy, with complete recovery. Moreover, two years after BCGitis, the patient is stable but remains under surveillance.\n\n\nConclusions\n\nWe describe in detail a very rare case of sepsis secondary to immunotherapy with BCG for a non-muscle invasive bladder cancer. Most frequently, patients have local manifestations, however systemic infection may also occur. The diagnosis can be challenging as other infections and malignancies must be first ruled out. Nonetheless, physicians ought to be familiar with bcgitis and upon diagnosis, prompt therapy must be started.\n\n\nConsent\n\nWritten informed consent for the publication of this case report and any associated images was obtained from the patient.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nBrausi M, Oddens J, Sylvester R, et al.: Side effects of Bacillus Calmette-Guérin (BCG) in the treatment of intermediate- and high-risk Ta, T1 papillary carcinoma of the bladder: results of the EORTC genito-urinary cancers group randomised phase 3 study comparing one-third dose with full dose and 1 year with 3 years of maintenance BCG. Eur Urol. 2014; 65(1): 69–76. PubMed Abstract \| Publisher Full Text\n\nGontero P, Bohle A, Malmstrom PU, et al.: The role of bacillus Calmette-Guérin in the treatment of non-muscle-invasive bladder cancer. Eur Urol. 2010; 57(3): 410–429. PubMed Abstract \| Publisher Full Text\n\nLamm DL: Efficacy and safety of bacille Calmette-Guérin immunotherapy in superficial bladder cancer. Clin Infect Dis. 2000; 31 Suppl 3: S86–S90. PubMed Abstract \| Publisher Full Text\n\nLencastre AR, R Pinheiro R: Immunotherapy-related genital BCGitis: Case report and a review of histopathology. J Cutan Pathol. 2019; 46(6): 462–465. PubMed Abstract \| Publisher Full Text\n\nPérez-Jacoiste Asín MA, Fernández-Ruiz M, López-Medrano F, et al.: Bacillus Calmette-Guérin (BCG) infection following intravesical BCG administration as adjunctive therapy for bladder cancer: incidence, risk factors, and outcome in a single-institution series and review of the literature. Medicine (Baltimore). 2014; 93(17): 236–254. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPommier JD, Ben Lasfar N, Van Grunderbeeck N, et al.: Complications following intravesical bacillus Calmette-Guerin treatment for bladder cancer: a case series of 22 patients. Infect Dis (Lond). 2015; 47(10): 725–731. PubMed Abstract \| Publisher Full Text" }	[ { "id": "57465", "date": "20 Dec 2019", "name": "André Lencastre", "expertise": [ "Reviewer Expertise Dermatology" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors report on a case of iatrogenic sepsis induced by BCG immunotherapy to treat bladder cancer. The final diagnosis was made with clinical history, ultrasound imagery and histology of the liver revealing a granulomatous hepatitis. No confirmation was possible testing for M. Bovis but the patient improved under specific treatment.\nThe overall structure and language of the document is elegantly built. I would suggest a more thorough review of cases of BCG induced sepsis or hematogenous spread to other organs, and its treatment.\nPlease review some of the language namely: 1) there are (virgula) spread all over the text; 2) there are units of concentration lacking or improperly written; 3) the phrase \"In a matter of days, the clinical outcome with resolution of the fever and normalized the inflammation markers at day seven from admission.\" has no verb; 4) (singular) is written somewhere in the text and makes little sense to me.\nI am kindly cited in the references but my name is not correct.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Partly\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [] }, { "id": "59867", "date": "24 Feb 2020", "name": "Marco Antonio Yamazaki-Nakashimada", "expertise": [ "Reviewer Expertise Clinical Immunology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors report a patient presenting an uncommon complication of BCG immunotherapy to treat bladder cancer. It is a very interesting case, with the peculiarity that the patient presented important radiological and histopathological liver involvement without jaundice or very elevated transaminases. Some authors suggest to reserve the term of BCGitis to local, regional BCG induced disease, and BCGosis to a distant or disseminated BCG induced disease, like in this case.\nI believe that the manuscript would be more valuable if the authors review all the cases of intravesical BCG treatment complicated with BCGosis involving the liver (Desmet et al., 20121, Bodurtha et al., 19742, Kaklamanos et al., 20113 etc.) and compare them to the present case. A table would be useful. Could you explain why corticosteroids were not used in this case since it presented in a severe fashion? Okano et al. (20194), reported a patient who was considered to have a hypersensitivity reaction to the liver after intravesical BCG who responded only to corticosteroids, and this study can be added and discussed.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Partly\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1399
https://f1000research.com/articles/8-788/v1	05 Jun 19	{ "type": "Case Report", "title": "Case Report: Case report: Pulmonary hemorrhage as a rare cause of lung ultrasound A/B-profile", "authors": [ "Mark E. Haaksma", "Esther J. Nossent", "Paul Elbers", "Pieter Roel Tuinman", "Esther J. Nossent", "Paul Elbers", "Pieter Roel Tuinman" ], "abstract": "When using lung ultrasound to determine the cause of acute respiratory failure, the BLUE protocol is often used. In a 65-year old patient, an A/B-profile was found, suggesting pneumonia, following the flowchart of this protocol. In this case, however, pulmonary hemorrhage confirmed by bronchoscopy was the final diagnosis. This case report outlines the importance of understanding the limitations of the BLUE protocol and that lung ultrasound findings should always be used in the context of the patient’s history and physical exam. In addition, pulmonary hemorrhage should be considered in patients with no clinical signs of pneumonia and/or presence of risk factors for lung bleeding as a rare cause of lung ultrasound A/B-profile.", "keywords": [ "Lung Ultrasound", "BLUE-Protocol", "Pulmonary Hemorrhage" ], "content": "Introduction\n\nOver the last few years, lung ultrasound has found routine use in critical care, even outperforming chest X-ray in detecting lung pathology1,2. Systematic approaches for its application are advised and have already been developed. For example, the BLUE protocol, developed by Lichtenstein et al., is often used to rapidly identify the cause of acute respiratory failure, with a claimed accuracy of 90.5%1. It uses interpretation of artifacts visible in the presence of pleural and/or pulmonary pathology on three distinct places. These include BLUE-1 on midclavicular line, BLUE-2 more caudally and lateral and the PLAPS-point (posterolateral alveolar and/or pleural syndrome point) on the dorsal side of the patient, continuing horizontally from the lower BLUE-point. The most frequently seen artifacts are lung-sliding and A-lines. The first being movement of the visceral pleura sliding against the parietal pleura and the second being a reverberation artifact of the pleural line. Their presence indicates unaffected lung tissue. B-lines are vertical hyperechogenic artifacts and are linked to fluid buildup in lung tissue. The presence of more than two B-lines indicates interstitial syndrome. Combining these artifacts on every point and following the schema, the physician is led to five distinct diagnoses, which are cardiogenic pulmonary edema, pulmonary embolism, pneumothorax, obstructive disease or pneumonia. The latter being characterized by unilateral B-lines resulting in an A/B-profile.\n\nIn the following article we present a case of pulmonary hemorrhage as a rare cause of lung ultrasound A/B-profile.\n\n\nCase presentation\n\nA 65-year old Caucasian male presented to the emergency ward with hemoptysis and respiratory distress. The patient’s history contained multiple pulmonary embolisms and subsequent chronic thromboembolic pulmonary hypertension, for which he was using acenocoumarol. Additionally, a non-specified form of pulmonary vasculitis was suspected, with concomitant treatment of prednisone. Bronchoscopy revealed pulmonary hemorrhage originating from the right pulmonary artery, which was coiled, and ICU admission followed. Two days after admission the patient was stable and extubated but showed signs of respiratory distress within the same day. Non-invasive ventilatory assistance failed and the patient was reintubated, while tranexamic acid (intravenous single dose, 1000 mg) was started as the cause was suspected to be a rebleed. Ultrasound examination according to the BLUE-protocol showed an A/B-profile (Figure 1a, b) and chest X-ray showed left-sided consolidation (Figure 1c). Subsequent bronchoscopy revealed a bleed originating from the left upper lobe (Figure 1d) which was successfully coiled.\n\n(a) A-profile (right lung). (b) B-profile with four distinguishable B-lines (left lung). Red dots, B-lines. (c) Chest X-ray with marked consolidation in the left lung. (d) Bronchoscopy of left main bronchus displaying bleeding from the upper lobe.\n\nThe following days were marked by respiratory instability and tracheal tube occlusion due to blood clots. One week after admission the patient developed atrial fibrillation, which was only temporarily relieved by cardiac resynchronization and amiodarone (300 mg intravenous loading dose in addition to 1200 mg/24 h intravenous for 1 day). This caused a further decline in cardiopulmonary function. A day later CT-angiography was performed and showed a fully occluded right bronchial artery and widespread occlusions in the left lung because of thrombosis. Due to the lack of additional therapeutic options, palliative care was started in line with the patient’s wishes, after which he passed away.\n\n\nDiscussion\n\nLung ultrasound has been demonstrated to be of high value in patients with respiratory symptoms. The BLUE-protocol is probably one of the most well-used protocols in these patients due to its ease of use and high accuracy. However, with all the enthusiasm around it, it is of vital importance to be aware of its limitations.\n\nThe protocol was originally developed for and tested in an emergency room population, although it is now frequently used in ICU patients, relying on a 90% diagnostic accuracy1. We don’t know, however, whether the same accuracy can be achieved in such a radically different patient population. ICU patients are frequently ventilated and could potentially be diagnosed with multiple pulmonary problems at the same time. In addition, pathology might be more subtle or different pathologies are present at the same time. To our knowledge, it is not clear how the BLUE-protocol performs in a setting with more than one underlying disease. Equally, the prevalence of included diagnoses differs between the ICU and the emergency room. This could potentially also influence its accuracy.\n\nFurthermore, the BLUE-protocol relies on simplification of reality by categorizing the etiology of dyspnea into five groups. This is also demonstrated by our case report, in which A/B profile was not caused by pneumonia but pulmonary hemorrhage. In line with this and the findings of our case report, we want to highlight how important it is to interpret the artifacts in the context of the patient’s history and physical exam. AB profile would normally suggest pneumonia as the underlying cause, but in the absence of infectious symptoms and parameters this seems unlikely. Pairing the profile with the patient’s recent history of pulmonary hemorrhage, this diagnosis also fits the picture. While this is only one example, it is not unlikely that other scenarios in which the seen artifact could be paired with more than one diagnosis can present itself.\n\nAside from ambiguity of profiles, another problem is the lack of certain diagnoses, such as ARDS (adult respiratory distress syndrome). In the ICU it is a frequently encountered problem, but has currently no place in the BLUE-protocol. Including more complex problems like these or even more rare diagnoses could presumably also impact its accuracy.\n\nLastly, we do not know what the optimal approach is, given that different available protocols have not been compared. Assessing only three points per hemithorax might not be sufficient and perhaps more comprehensive imaging with more areas investigated are necessary3,4. The optimal balance of areas screened vs. time spent on an exam is yet to be determined. There are also no strict guidelines in regards to choice of probe and their settings during examination. This also might influence the quality of images and hence the interpretation of artifacts. In the future, we therefore hope to see more elaborate guidelines that encompass detailed information for image acquisition. Inclusion of more diagnoses with perhaps with more advanced lung ultrasound modalities, such as Doppler or 3D imaging, would help to develop lung ultrasound to an even more sophisticated diagnostic tool.\n\n\nConclusion\n\nThis case report outlines the importance of understanding the limitations of the BLUE-protocol and that lung ultrasound should always be used in the context of the patient’s history and physical exam. Also, pulmonary hemorrhage should be considered in patients with no clinical signs of pneumonia and/or presence of risk factors for lung bleeding as a rare cause of lung ultrasound A/B-profile.\n\n\nData availability\n\nNo data are associated with this article.\n\n\nConsent\n\nWritten informed consent for publication of clinical details and clinical images was obtained from the patient’s family.", "appendix": "Grant information\n\nThe author(s) declare that no grants were involved in supporting this work.\n\n\nReferences\n\nLichtenstein DA, Mezière GA: Relevance of lung ultrasound in the diagnosis of acute respiratory failure: the BLUE protocol. Chest. 2008; 134(1): 117–25. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWinkler MH, Touw HR, van de Ven PM, et al.: Diagnostic Accuracy of Chest Radiograph, and When Concomitantly Studied Lung Ultrasound, in Critically Ill Patients With Respiratory Symptoms: A Systematic Review and Meta-Analysis. Crit Care Med. 2018; 46(7): e707–e714. PubMed Abstract \| Publisher Full Text\n\nJambrik Z, Monti S, Coppola V, et al.: Usefulness of ultrasound lung comets as a nonradiologic sign of extravascular lung water. Am J Cardiol. 2004; 93(10): 1265–1270. PubMed Abstract \| Publisher Full Text\n\nVolpicelli G, Mussa A, Garofalo G, et al.: Bedside lung ultrasound in the assessment of alveolar-interstitial syndrome. Am J Emerg Med. 2006; 24(6): 689–696. PubMed Abstract \| Publisher Full Text" }	[ { "id": "49573", "date": "21 Jun 2019", "name": "Belaid Bouhemad", "expertise": [ "Reviewer Expertise Ultrasound", "ICU" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAuthors described limitations of the «BLUE protocol» about one case of hemoptysis. The authors should be commended for this task. The case report is well written, “English” seems to me very good. I have only 2 comments to add to the discussion part:\nBlue protocol is not for ICU, we agree, and Lung US ICU protocols or guidelines are not available. But review written by experts are available. Please quote Bouhemad Anesthesiology, 2015 Feb;122(2):437-471, and Mojoli, Am J Respir Crit Care Med, 2019 Mar 15;199(6):701-7142. Moreover, in these papers, the issues of the needed “examinations points” (at least 6 X right/left) and probe are discussed.\n\nIn the conclusion, authors should precise that lung ultrasound \"as all imaging technique should be considered in the patient’s….etc\".\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Partly\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [ { "c_id": "4779", "date": "30 Jul 2019", "name": "Mark Haaksma", "role": "Author Response", "response": "Dear Dr. Bouhemad, Thank you for your time and expertise. We have adjusted the manuscript according to you feedback. The mentioned reviews are indeed an excellent addition to this article. We also agree that all imaging modalities have to be seen in the context of the patient." } ] }, { "id": "49575", "date": "27 Jun 2019", "name": "Daniel A. Lichtenstein", "expertise": [ "Reviewer Expertise Critical care medicine", "critical ultrasound", "lung ultrasound." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nReviewing this interesting case report was a fine opportunity to make basic reminders on the BLUE-protocol. We hope this will allow the readers to better understand how to use it.\n\nWe briefly correct minor details. The authors speak of BLUE-1 and BLUE-2. Following the lung ultrasound nomenclature, they should speak of upper and lower BLUE-points1. Four lines before Conclusions, minor typo (with perhaps with).\n\nThis being said, the good point is the illustration of a rare cause of dyspnea (pulmonary hemorrhage) seen through the BLUE-protocol (and generating an A/B-profile). This is important to know for future uses. We now have to comment numerous other points (focusing on the most important for shortening our analysis).\n\nThe article begins softly by a simple case report of a rare situation, and then increasingly challenges several areas of the BLUE-protocol, not specifically related to the topic on focus (the non infectious A/B-profile). While respecting and understanding the approach of the authors, our review will explain again what is the BLUE-protocol - a precious reminder for those who did not read this article.\n\nThis case report is presented as a limitation of the BLUE-protocol. It should, on the contrary, be understood as a perfect illustration of its spirit. May we remind that the BLUE-protocol is a clinical protocol. First of all (this is specified in its title), it includes the critical notion that the patient is in respiratory failure. If the BLUE-protocol was independent from this detail, we would all have severe asthma or exacerbated COPD, that is, the nude profile. Second, the BLUE-protocol is only the ultrasound part of a clinical approach to the patient. The BLUE-protocol is always performed after having received simple data of a simple history and a simple physical examination (see below for this case report).\n\nLater, the authors consider that “the BLUE-protocol relies on [over] simplification”, with only five groups of diseases. May we remind that it has considered 97% of the patients (adults, seen in the ER, all admitted to the ICU)? We also remind that the majority of diagnostic mistakes (and dramas) include these frequent causes of dyspnea, this is why the BLUE-protocol focuses on the real, daily life. By the way, the 3% of other patients had countless causes (including this interesting case report). Creating a decision tree with all causes of dyspnea would generate a decision tree beyond any description in terms of complexity. Anyway, an answer to the concern of the authors can be consulted in our latest textbook, the E-BLUE-protocol, that is, an extended BLUE-protocol2. They will find more developments of each profile by using tools from simple (echocardiography e.g.) to complex and invasive procedures.\nFor each of the usual diseases, the reader benefits from various values of accuracies, which indicates that the BLUE-protocol is not designed for giving but suggesting (more or less highly) diagnoses.\nTherefore, we are here in a caricatural situation where the doctor has immediately detected a particular setting and is therefore invited to promptly use additional tests (as the authors rightly did in this case report). Here for instance, the history immediately points on a very particular situation (a long history of pulmonary embolism), and the physical examination immediately detects a major sign (hemoptysis) (and, we guess, absence of fever). The BLUE-protocol can of course be performed, but for building a diagnosis, such patients are immediately excluded from its thought process, and managed independently.\n\nThis being said, we will comment on several points through the article:\nLung sliding and A-lines are “unaffected lung tissue”. We rather speak of unaffected lung surface. We remind that the A-profile (anterior bilateral lung sliding plus predominant A-lines) can be seen in pulmonary embolism (generating the A-DVT-profile), pure posterior pneumoniae (A-V-PLAPS-profile) and COPD and/or asthma (nude profile).\n\nThe authors consider that three points are too limited. Yet they cautiously use the word “perhaps”, and we invite them to test the BLUE-protocol with more points. We have done this comparison since three decades and did not find a significant change in the accuracies. We never found time for submitting our data, and strongly encourage the authors to making this comparison, and publish it.\n\nB-lines are “linked” to fluid buildup. This is one of four possibilites only. Stricto sensu, the B-line indicates fluid surrounded by air (in a ratio of 1%-99% roughly). This fluid can be interstitial; it can be transudative in the case of hydrostatic pulmonary edema (B-profile usually), or exudative in the case of pneumonia/ARDS (B’-profile in many cases). The fluid can be intra-cellular; normal in the case of a simple lung fissure (isolated B-line), or pathologic on the case of chronic interstitial disease (usually lung rockets).\n\n“The presence of more than two B-lines” is not sufficient. The diagnosis of interstitial syndrome is based, not on a number, but on a concentration of B-lines. The problem is (very simply) solved if you add “between two ribs”, that is a distance of roughly 2 cm, that is, 3 B-lines every 2 cm.\n\n“Unilateral B-lines resulting in an A/B-profile”. Please replace “B-lines” by “lung rockets” and it will be correct. The B-line is not a disease, we almost all have B-lines. Reminder, more than two B-lines between 2 ribs are called lung rockets.\n\nThe authors write: “it is not clear how the BLUE-protocol performs in a setting with more than one underlying disease”. If they raise the issue of mixt diseases, these cases are integrated in the E-BLUE-protocol. In the native article, such patients were excluded for obvious methodological reasons (this affected roughly 1% of the patients). If a patient has two conditions potentially generating dyspnea, which gold standard will be able to determine the respective part of both diseases? It can be 50-50%, it can also be 51-49%, up to, potentially, 99-1%.\n\nThe authors write that ARDS has “currently no place in the BLUE-protocol”. This is wrong. The lecture of the native article shows that ARDS is considered as a pneumonia (severe, extensive, deadly, but just a pneumonia), which makes sense in the large majority of cases, again for simplifying a daily situation. The four profiles of pneumonia (B’, AB, C and AVPLAPS-profile) are found in the majority of cases of ARDS.\n\nThe authors “don’t know whether the same accuracy can be achieved in such a radically different patient population” (emergency room versus ICU). First, some diagnoses can perfectly be achieved in ICU patients. A pneumothorax still is a pneumothorax. A major pleural effusion is still a pleural effusion. An increase of lung rockets (from septal rockets to glass rockets, e.g.) indicates a worsening (and many other examples). Yet the authors are partially right, these are two different groups. The answer is simple. The BLUE-protocol is reserved for the initial diagnosis (usually in the ER). For ICU patients, submitted to mechanical ventilation and severe diseases with fast changes, the CEURF (our academic teaching centre) has designed a specific approach, the Pink-protocol. These patients have no dyspnea and no cyanosis (hence this label), but they have no more lung function. The technique is more comprehensive that the BLUE-protocol (more points of investigation), and the interpretation is, obviously, more complex. Obstructive atelectases, e.g., much more frequent in this group, are integrated within the Pink-protocol.\n\nAccording to the authors, there is “no strict guidelines in regard to choice of probe”. We just recall the probe which we have used through all our publications on whole body ultrasound since our first success, submitted since 19913. We use it for assessing abdominal disorders, deep venous thromboses, and all our procedures (including subclavian cannulation). We do not prevent the community to use classical probes, we just see the drawbacks of this approach: cost (buying probes), time (changing probes), infectious issues (cleaning probes - we are unable to avoid asepsis errors when using more than one probe) and last, a more complicated approach to the subtle science of lung ultrasound. Our probe is unique because of its logical ergonomy (allowing all accesses), its range allowing to detect both superficial and deep details (using the criteria 2 and 4 of the B-line), and its clear ability to visualise B-lines (using criteria 5 and 7) (Fig. 1).\n\nIn the Figure 1B, we do not see any B-line. Or, we just guess some, ill-defined, in the fog - and our 3-decade experience helped us a lot. How about novice users? The criterion N°5 of the B-line: “well-defined” (laser like), and the criterion N°7: “hyperechoic” (that is, white like the pleural line) are missing. This is a severe concern. Look our world (available since at least 1982), and please compare (Fig. 1). The result is a hindrance to an optimal development of lung ultrasound, and the need for complicated solutions. With a simple machine and probe, B-lines need no sophistication, no artificial intelligence, no automatic detection, no 3-D nor Doppler imaging.\n\nSuggested solutions:\n\nPlease try to disactivate all filters of your modern machine, maybe you will succeed to have a convincing image. The manufacturer must not only sell you the unit, they must also find where are all those filter which, to our opinion, have no utility in critical ultrasound but just resulting in such images.\nThe authors argue for Doppler and 3D imaging. Here, our answer is simple. Since three decades, we promise to use these sophisticated techniques as soon as we will feel limited by the simple and standardized approach provided by CEURF. This time has still not come. When we see the provided image quality, we understand why people develop sophisticated solutions. The physicians who use suitable equipment will, likely, not need these sophistication's. On purpose, we insist on this point.\n\nWe take the advantage of this providential tribune for pointing out that the value of “100%” likely does not exist in medicine (apart from privileged exceptions, such as the lung point). In the native article, the A/B-profile was 100% specific to pneumonia. This beautiful performance was of course due to a weak number (12 of 300 studied patients had the A/B-profile). It is completely intuitive that multiplying the numbers will make appear exceptional situations. If we make a study of 1000 cases of AB-profile, rare diseases will be seen sometimes. We will see maybe 10 or 20 noninfectious causes, not a lot more we guess. Therefore, the specificity of the A/B-profile would become 99%, maybe 98%, not less. This is the basic statistical art.\n\nWe are of course sorry about the outcome of this patient. Some academicians may argue that if a tool is used in an exceptional setting, makes anyway a diagnosis, but does not help the patient from a sad outcome, this tool was eventually of limited value here. We respect life too much for just thinking this.\n\nConclusions:\nThis review was a nice opportunity for a pedagogic exercise: reminding what is the BLUE-protocol. If we were asked to summarize our review in a few words, we would write: “Please pilot the BLUE-protocol. Be a doctor... and it will give its best”. We congratulate the authors to use this attractive part of lung ultrasound, and we encourage them for any further investigations.\nFigure 1: (Figure 24 (middle image)) B-lines, as they were described in the CEURF nomenclature with their seven criteria. Three are constant (comet-tail artifact, arising from the pleural line, moving in sync with lung sliding). Four are almost constant (long, well-defined, erasing A-lines, hyperechoic). Here, typical lung rockets. From a 1992 Hitachi 405 and its microconvex probe. Note that in the figure under link, the terms B7 and B3 lines were long replaced by the much more speaking septal rockets and glass rockets, respectively.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [ { "c_id": "4780", "date": "30 Jul 2019", "name": "Mark Haaksma", "role": "Author Response", "response": "Dear Dr. Lichtenstein, Thank you for your time and expertise. We have adjusted the manuscript according to you feedback and would also like to respond to your comment. We would like to point out that this article was in no way meant to directly critique the BLUE-protocol. We ourselves are avid users of said protocol and ultrasound in general. The enthusiasm around it is, rightfully so, rapidly spreading. With it however, we have noticed that eagerness to quickly implement LUS in daily ICU practice comes with a lack of caution. Therefore, this article is meant as a reminder of certain caveats, which do not necessarily stem from the protocol but its implementation. We agree that in this instance the cause for respiratory failure was relatively clear and that application of the BLUE protocol was not called for. However, we believe with this clear-cut situation we underline our points given in the discussion and above. Point being that the high accuracy of the protocol should not lead physicians to a misplaced sense of security in more dubious cases. This again, is more as consequence of its implementation and not the protocol itself. In regard to image and probes we agree that the presented microconvex probe probably has various benefits over other probes that are frequently used. However, exactly the fact that various probes are being used is of great importance. Physicians should, yet sometimes might not be, aware of this. In our daily practice, we also see that there are differences in image quality depending on probes used. While we agree that “simple” LUS is sufficient for B-line assessment, we still believe that in the spirit of scientific curiosity we are obliged to research whether arising modalities such as 3D and doppler imaging might contribute to it. Perhaps new possibilities will present itself in this regard.Nevertheless, we want to thank you for your elaborate comment. It is without a doubt a good reminder and lesson of the BLUE-protocol for all readers." } ] } ]	1	https://f1000research.com/articles/8-788
https://f1000research.com/articles/8-1372/v1	07 Aug 19	{ "type": "Research Article", "title": "Determinants of drug entry into the developing brain", "authors": [ "Liam Koehn", "Mark D. Habgood", "Yifan Huang", "Katarzyna M Dziegielewska", "Norman R. Saunders", "Liam Koehn", "Mark D. Habgood", "Yifan Huang", "Katarzyna M Dziegielewska" ], "abstract": "Background: A major concern for clinicians in prescribing medications to pregnant women and neonates is the possibility that drugs might have damaging effects, particularly on long-term brain development. Current understanding of drug permeability at placental and blood-brain barriers during development is poor. In adults, ABC transporters limit many drugs from entering the brain; however, little is known about their function during development. Methods: The transfer of clinically relevant doses of paracetamol (acetaminophen), digoxin and cimetidine into the brain and cerebrospinal fluid (CSF) was estimated using radiolabelled drugs in Sprague Dawley rats at three developmental stages: E19, P4 and adult. Drugs were applied intraperitoneally either acutely or following chronic exposure (for five days). Entry into brain, CSF and transfer across the placenta was measured and compared to three markers (L-glucose, sucrose, glycerol) that cross barriers by “passive diffusion”. The expression of ABC transporters in the brain, choroid plexus and placenta was estimated using RT-qPCR. Results: All three drugs entered the developing brain and CSF in higher amounts than the adult brain and CSF. Comparisons with “passive” permeability markers suggested that this might be due to age-related differences in the functional capacity of ABC-efflux mechanisms. In adult animals, chronic treatment reduced digoxin (12% to 5%, p<0.01) and paracetamol (30% to 21%, p<0.05) entry compared to acute treatment, with the decrease in digoxin entry correlating with up-regulation of efflux transporter abcb1a (PGP). In fetal and newborn animals, no gene up-regulation or transfer decreases were observed. Instead, chronic paracetamol treatment resulted in increased transfer into the fetal brain (66% to 104%, p<0.001). Conclusions: These results suggest that the developing brain may be more at risk from acute drug exposure than the adult brain due to reduced efflux capacity and at greater risk from chronic treatment due to a lack of efflux mechanism regulatory capacity.", "keywords": [ "ABC transporter", "Blood-brain barrier", "fetus", "neonate", "cerebrospinal fluid", "placenta", "permeability" ], "content": "Abbreviations\n\nABC, ATP-binding cassette; BCRP, breast cancer resistance protein; CSF, cerebrospinal fluid; DPM, disintegrations per minute; E, embryonic (note that by longstanding convention all gestational ages in rodents are referred to as embryonic, but in this study E19 is a fetal stage); i.p., intraperitoneal; i.v., intravenous; MRP, Multidrug resistance-associated protein; P, postnatal; PGP, P-glycoprotein; RT-qPCR, Real time quantitative polymerase chain reaction; SD, standard deviation; µCi, micro Curie.\n\n\nIntroduction\n\nThe mechanisms that prevent or limit entry of drugs and toxins into the adult brain are reasonably well known. For water-soluble molecules, intercellular transfer is largely prevented by tight-junctions (Johansson et al., 2008; Saunders et al., 2008; Saunders et al., 2018). For lipid soluble compounds, transcellular transfer of many compounds is limited by efflux transporter mechanisms (Mandal et al., 2017). Members of the ATP-binding cassette (ABC) transporter family, known to be located in the various brain barrier interfaces (Roberts et al., 2008; Saidijam et al., 2018; Strazielle & Ghersi-Egea, 2015), are major contributors to this protection. They are the main reason why it has proved so difficult to develop new drugs for neurological and neuropsychiatric conditions. It has been estimated that 98% of drugs developed by pharmaceutical companies for such conditions fail to enter the brain in therapeutically useful amounts (Pardridge, 2002).\n\nIn contrast to the knowledge about the adult brain, little is known about the presence and functional activity of ABC transporters in the developing brain (Strazielle & Ghersi-Egea, 2015; Saunders et al., 2019). This information is essential for understanding the likelihood of a drug to enter the brain directly from the fetal circulation (once the substance has crossed the placenta) or from the circulation of a newborn (especially pre-term), who lacks placental protection. In the clinic, drugs are administered to pregnant women and newborns for a range of conditions and over 1200 drugs have been prescribed during pregnancy and lactation (Briggs et al., 2017). Evidence of potential harms, except in a few cases, is unclear. The scale of the problem is illustrated by international surveys showing that in all countries studied the proportion of pregnant women who take medications during pregnancy is high (Wyszynski & Shields, 2016). While the clinical application of these drugs in adults is supported by evidence from clinical trials, such trials have not been conducted in pregnant women and neonates because of obvious ethical concerns (Lyerly et al., 2008). In the absence of these data, doctors have to rely on their experience of observed side effects following the application of drugs to these patient populations. However, in the case of the central nervous system (CNS) the harmful effects may not manifest themselves until much later in baby’s development, making them difficult to track. While controlled clinical trials in these patient populations may remain difficult, enhancing the understanding of barrier permeability and ABC transporter functionality at different developmental stages would provide additional evidence to aid clinicians.\n\nAnimal studies on the entry of drugs into the developing brain following administration to pregnant animals or newborns are also very scarce. The few studies that have been published on pregnant rodents have been reviewed in Saunders et al. (2019) and will be further considered in the Discussion. To determine the relative contributions of the placental and brain barriers in drug protection of the developing brain, individual measurements in fetal blood and cerebrospinal fluid (CSF) are required. Access of any molecule into the CNS is determined by: (i) their physicochemical properties, such as molecular size and lipid solubility; (ii) biological properties (facilitated transfer by influx mechanism, e.g. glucose, or restricted transfer by efflux mechanisms such as ABC transporters); and (iii) physiological properties of brain barriers that are developmentally regulated (e.g. CSF secretion). The actual level that a molecule reaches in brain and CSF is also influenced by the turnover of CSF (“sink action”, Davson, 1967), which is much less in the developing brain (Saunders, 1992). Transfer of lipid insoluble (hydrophilic) molecules that are passively transferring (i.e. not transported) from blood into the brain and CSF is determined by their molecular size at any stage of brain development (Dziegielewska et al., 1979; Habgood et al., 1993). For lipid soluble (hydrophobic) molecules, their permeability is dependent not only on their degree of lipid solubility (octonol/water coefficient or logDoctonol, Davson, 1967; Rapoport et al., 1979) but also on their specificity, if any, for individual ABC transporters that play an active role in molecular exclusion at the barrier. It is this last function that is little understood during brain development but is critical to understand their possible limitation of drug entry into the brain at different stages of its maturation.\n\nA key biological problem, which has implications for understanding potential deleterious effects of drugs administered to pregnant women or new-born infants, is the stage of brain development when the efflux transporters appear and when they become functionally effective. It is not necessarily the case that once a transporter is present it will show the same level of activity throughout development; this might increase or decrease at different times (see Discussion and Ek et al., 2010; Møllgård et al., 2017).\n\nThe present paper describes experiments using three index drugs that are given to pregnant women and/or neonates (paracetamol, digoxin and cimetidine). Two of the drugs selected (digoxin and cimetidine) are used for their peripheral therapeutic effects (cardiovascular and alimentary systems respectively) but there is some evidence that they do enter the brain to a limited extent in adult animals (digoxin: Liu et al., 2014; Mayer et al., 1997; Taskar et al., 2017 and cimetidine: Kodaira et al., 2011). There are a few reports of entry of digoxin and cimetidine into the brains in fetal rodents (see Discussion). The third drug, paracetamol, is the most widely used drug during pregnancy (Werler et al., 2005). It is the only analgesic that is regarded as “safe” in infants (Australian Medicines Handbook, 2019; World Health Organisation, 2012). These three drugs are thought to be substrates for different ABC transporters. There are studies linking PGP (abcb1) to digoxin transfer (Mayer et al., 1996; Smit et al., 1999) and BCRP (abcg2) to cimetidine transfer (Liu et al., 2007; Staud et al., 2006). It is unclear from the literature which ABC transporter(s) paracetamol is a substrate for, although it has been suggested that MRP3 (abcc3) may limit paracetamol entry via its glucuronidated metabolite (Manautou et al., 2005). Paracetamol’s other glutathione and sulphate metabolites may be substrates for BCRP (abcg2) or other MRP (abcc) transporters. All three drugs are available in a radiolabelled form. The use of radiolabelled drugs is essential for detection of a drug in the very small volumes of plasma and CSF available from fetal and neonatal rodents.\n\nThe results showed that there are clear age-dependent differences in the entry of the three drugs (paracetamol, digoxin, cimetidine) into brain and CSF. In addition, over the course of chronic exposure, drug entry was reduced only after a certain stage of brain maturation. Some of these differences appear to be accounted for by changes in the expression levels of individual ABC transporters. The study provides a basis for future comprehensive research into the entry of a wide range of drugs into the developing brain.\n\n\nMethods\n\nAll procedures involving animals were approved by the University of Melbourne Animal Ethics Committee (Ethics Approval AEC: 1714344.1) and conducted in compliance with Australian National Health and Medical Research Guidelines. All efforts were made to ameliorate any suffering of animals. They were handled by experienced researchers in such a way as to minimise stress prior to being anaesthetised. All animals were assessed as healthy prior to commencement of experiments. Animals were monitored prior to and following every injection ensuring there was no abnormalities in weight (>15%), appearance (wounds, fur) or behaviour (vocalisation, respiration, movements).\n\nAnimals. Sprague Dawley rats were supplied by the University of Melbourne Biological Research Facility and housed in groups of 2–4 (adult) or full litters per cage (25cm x 35cm x 25cm on Breeders Choice paper bedding, made from 99% recycled paper; it is biodegradable with no added chemicals), on a 12 h light/dark cycle with ad libitum access to food (dry pellets of a fixed formulation diet for laboratory rats and mice fortified with vitamins and minerals to meet the requirements of breeding animals after the diet is autoclaved or irradiated, supplied by Speciality Feeds, Western Australia) and water. Age groups investigated (at treatment completion) were time mated pregnant females at E19 (350-400g body weight), postnatal pups at P4 and non-pregnant female adults (175-230g body weight). E19 was chosen because this is a stage of development when adequate volumes of blood and CSF can be obtained for analysis from fetal rats without pooling (Dziegielewska et al., 1981) and individual pups can be injected intraperitoneally while still inside the uterine horn and kept viable for periods of time. P4 was chosen because its stage of brain development is similar to that of very prematurely born but viable human infants of 22–24 weeks gestation (Clancy et al., 2001). The numbers of animals (n) used for each experiment are indicated in Table 1 and Table 2. Animal numbers were based on previous experience of such experiments and were the minimum number required to detect a significant difference between groups at P <0.05. Animals were selected for treatment groups to ensure weights were statistically similar between direct comparisons.\n\nThe n numbers listed in brackets are where the numbers of CSF measurements were different from the number of brain measurements. No animal died before the termination of an experiment. E19 and P4 pups were littermates and included both sexes. The differences in pup numbers between groups is due to natural variation in litter sizes. P4 and adult experiments were terminated 30 minutes after tracer administration, whereas the E19 embryo experiments were terminated between 30 minutes and 2.5 hours post-injection due to the time required to sequentially sample each embryo. Acute: experiments involved a single injection of unlabelled drug mixed with a tracer amount of radiolabelled drug. Chronic: animals received multiple drug injections of unlabelled drug over a period of five days (see Methods) with the final injection including the radiolabelled tracer. Adults were non-pregnant females.\n\nAll experiments were acute (single injection including the tracer) with samples collected at 30 minutes post-injection, except for placental transfer experiments where blood samples were collected between 30 minutes to 105 minutes post injection. No animal died before the termination of an experiment. E19 and P4 pups were littermates and included both sexes. Glycerol E19 experiments were conducted in litters from two separate pregnant females. Adults were non-pregnant females.\n\nDrug doses. Drug doses were selected based on use in clinical practice (Australian Medicines Handbook, 2019) and adjusted for body weight. Cimetidine (C4522, Sigma-Aldrich) was applied at 11mg/Kg per dose, digoxin (D6003, Sigma-Aldrich) at 30μg/Kg per dose and paracetamol (acetaminophen ≥99.0%, Sigma-Aldrich) at 15mg/Kg per dose.\n\nCimetidine and paracetamol were dissolved in sterile 0.9% sodium chloride solution.\n\nDigoxin was dissolved in ethanol before dilution in sterile 0.9% sodium chloride solution for injection, with final injectate ethanol concentration <5%. Radiolabelled drugs and hydrophilic markers used are listed in Table 3.\n\nExperimental procedure. In acute drug experiments, a single dose was injected that included traces of [3H]-labelled drug (20µCi adults, 2µCi newborns). In chronic experiments, doses of unlabelled drug were given twice daily (09.00 and 17.00h) for four days before a final injection on the 5th day identical to acute experiments. In passive marker experiments, all dosing was acute and contained only the [3H]- or [14C]-labelled compound (20µCi adults, 2µCi newborns).\n\nIn all experiments involving postnatal animals, injections were intraperitoneal (i.p.). In pregnant animals for fetal drug entry studies, the final injection of the radiolabelled marker was given intravenously (i.v.), as the peritoneal cavity required opening prior to the sampling period to gain access to the fetuses. For passive markers (L-glucose, sucrose and glycerol), fetal animals were individually injected i.p. while still within the uterine horn. All experiments took place between 09.00 and 15.00h. All pregnancy studies were completed with a single treatment group on one day. Postnatal groups were conducted with a single treatment group on one day or with the chronic group completed prior to the acute group.\n\nSample collection. Samples were collected 30 minutes after the final injection. This duration was chosen partly to limit any potential metabolism of the drugs and markers used but also to allow enough time for i.p. injected markers to reach the blood stream and, from there, to access the brain parenchyma across brain barriers, as well as to limit the CSF sink effect that could influence drug levels (Bito et al., 1966). Thirty minutes is also within the half-life of these drugs (Adedoyin et al., 1987; Harrison & Gibaldi, 1976; Lin & Levy, 1963). The other reason a 30-minute duration was chosen is that in the case of exposed fetal rat pups, it is difficult to maintain them in a reasonable physiological state for long periods due to deterioration of placental perfusion. Therefore, for comparative purposes, all studies were conducted at the same time.\n\nFor adult and newborn (P4) experiments, animals were terminally anaesthetized using inhaled isofluorane (IsoFlo 100% w/w, Abbott Laboratories). Blood samples were collected from the right cardiac ventricle, CSF from the cisterna magna (Habgood et al., 1992) and cortical segments of brain tissue dorsal to the ventricle of the frontal/parietal lobes, as previously described (Koehn et al., 2019). Blood and CSF sampling from the P4 animals was terminal and would thus not have affected the physiological state of the animals during the experiment.\n\nIn pregnancy experiments, animals were anaesthetised i.p. with 25% w/v urethane, Sigma, 1ml per 100g body weight, dose). Animals were placed on a 33°C heating plate in a supine position and an endotracheal catheter inserted to maintain a clear airway. The femoral artery and vein of the left hindlimb were cannulated to provide access for collecting arterial blood samples and intravenous (i.v.) injection of the radio-labelled drug or marker solution; the cannula was flushed with 1ml of heparinized (Hospira Inc, five units per ml) saline. All injections were made by slow infusion. Starting at 30 minutes after injection blood, CSF and brain samples (following terminal exsanguination) were collected serially from each fetus. At the time of each fetus sampling, the state of the placental circulation to the fetus was assessed from the colour of the umbilical blood vessels (pink veins indicate reasonable oxygenation). Time matched blood samples (200μl) were collected from the maternal arterial cannula to establish levels of radioactivity in the maternal circulation at the time of each fetal sampling. Maternal blood volume was maintained by intraarterial injection of equivalent volumes of heparinized sodium chloride solution, which also served to maintain the patency of the cannula. Fetuses were sampled over a period of 30 minutes - 2.5-hours post-injection, the time during which placental circulation to the fetuses remained adequate.\n\nSample preparation. Samples were processed immediately after collection. Blood was centrifuged at 7000rpm for five minutes and the plasma separated. CSF samples were examined microscopically for traces of red blood cells and discarded if contaminated (Habgood et al., 1992). In all experiments, the radioactivity in the injectate was also measured to confirm the uniformity of the injected material. All samples were weighed and transferred into scintillation vials. Soluene350 (0.5ml, PerkinElmer) was added to the brain samples. After being incubated at 36°C overnight to allow the tissue to solubilize, two drops of glacial acetic acid (Sigma) were added to neutralize the strongly alkaline Soluene350. All samples were then mixed with 5ml of scintillation fluid (Emulsifier-safe, PerkinElmer) before being transferred into the liquid scintillation counter (Tri-Carb 4910 TR, PerkinElmer) to count radioactivity disintegrations per minute (DPM) over five minutes each with luminescence correction on. Blank vials containing the same tissues without radioactivity were also counted alongside the samples to establish the level of background counts, which were always subtracted from the radioactivity counts of the corresponding experimental samples.\n\nSample analysis. Background-corrected DPM data from the liquid scintillation counter were normalized to the weight of the samples and expressed as DPM per µl or µg of sample. Ratios denoting brain and CSF transfer for all animals (Equation 1) and additional placental transfer between the fetus and the mother for fetal animals (Equation 2) were obtained:\n\nEquation 1\n\nBrainorCSFtransfer=BrainorCSFDPM/μlplasmaDPM/μl×100%\n\nEquation 2\n\nPlacentaltransfer=FetalplasmaDPM/μlMaternalplasmaDPM/μl×100%\n\nSample collection. For RT-qPCR studies, a new set of animals underwent the same experimental procedures as those for the drug entry experiments (above), except that no radioactive tracers were included. Before tissue collection, all surgical instruments were cleaned with RNaseZAP (Thermo Fisher Scientific) to destroy any RNases. For the E19 group, the tissue collected from each pup was brain (see Drug Entry section) and placenta. The small lateral ventricular choroid plexuses were also collected and pooled from all pups into one sample. Tissue collected from P4 and adults (brain and choroid plexus) were processed individually. The samples were placed into sterile cryogenic vials, snap frozen in liquid nitrogen and then transferred into a -80°C freezer for storage.\n\nRNA extraction and reverse transcription. RNA was extracted using RNeasy Plus Mini Kits (QIAGEN, Cat no 74134) for the brain samples in the postnatal animals and E19 placenta samples according to manufacturer’s specifications. For the fetal brain samples and all choroid plexuses, RNA was extracted using RNeasy Micro Kits (QIAGEN, Cat no 74004) according to manufacturer’s specifications. RNA purity and quantity were assessed using a Nano-drop (ND-1000 UV-VIS spectrophotometer, Thermo Scientific) and the concentration of all samples standardized in nuclease-free water (<300ng/μl). Each RNA sample was then converted to cDNA using the High Capacity RNA-cDNA kits (Applied Biosystems, Cat no 4387406) containing 9µl of RNA in Nuclease-free water, 10µl 2x reverse transcriptase buffer mix (RT buffer) and 1µl 20x RT enzyme mix, making a total volume of 20µL. RNA was converted to cDNA using a thermocycler (Veriti 96 Well, Applied Biosystems) set at 35°C for 60 minutes, followed by 95°C for five minutes and then incubated at 4°C until collection. The more stable cDNA was stored at -80°C until ready to analyse with the RT-qPCR system (QuantStudio 6 Flex, Applied Biosystems).\n\nRT-qPCR. Regulation of gene expression was measured using RT-qPCR and amplification detected with SYBR Green fluorescence. Primers listed in Table 4 were made at 800nM, except for abcc1, which was at 400nM to avoid reads in no-template control (NTC) wells. PCR efficiencies were validated. The final 10µl reaction volume contained 2µl 1/10 diluted cDNA and 8µL SYBR Green master mix (5µl Rt2 SYBR Green ROX qPCR master mix, [QIAGEN, Cat no 330529], 1µl each forward and reverse primer, 1µl RNase free water). No-template controls (NTCs) were prepared and analysed alongside the cDNA triplicates with every run. NTCs produced no signal, or the occasional negligible signal (> 38Ct). Plates were evaluated using the Applied Biosystems Quantstudio6 Flex machine: two minutes at 50°C, 10 minutes at 95°C, 40 amplification cycles for 30 seconds at 95°C and one minute at 60°C. Transcript counts were compared to a housekeeper peptidylprolyl isomerase B (ppib) using the equation 2-ΔCt.\n\nGroup permeability data (brain/plasma and CSF/plasma concentration ratios) for the acute and chronic conditions at all ages are presented as mean ± standard deviation (SD). Microsoft Excel 2011 was used for statistical analysis. Statistical differences between the acute and chronic groups for each drug were determined by unpaired, two-tailed Student’s t-test with p<0.05 accepted as significant and F-tests to confirm equal variance.\n\n\nResults\n\nThe present study investigated the degree of protection that is present at different developmental stages to prevent or limit the transfer of drugs from the circulation into the brain. Rats were injected with either a single dose (acute experiments) or multiple doses (chronic experiments) of one of three drugs (digoxin, cimetidine or paracetamol) and analysed at E19, P4 and adult (see Methods). Results from in vivo drug studies were compared with gene expression data to establish which ABC transporters changed their expression following chronic exposure, in order to see if they correlated with observed changes in drug entry results. The transporters studied were: abcc1-5 (MRP1-5), abcg2 (BCRP) and abcb1a/abcb1b (p-glycoprotein, PGP, which in the rodent has two isoforms). For quality control, the radioactivity counts estimated in different compartments in all experiments are shown in Table 5. The results are then displayed as brain/plasma and CSF/plasma concentration ratios (Table 6). This representation is a convention used in many blood-brain barrier experiments (Davson & Segal, 1996); this is because the entry of a marker will depend largely on its amount in circulating blood, which inevitably varies to some extent between experiments (as can be seen from the standard deviations in Table 5). The values in Table 6 and in Figure 1–Figure 3 for each drug represent a measure of “apparent permeability” rather than actual “permeability” because of the influence of the secretion of CSF, which is much lower in the developing brain (Saunders, 1992). Changes in expression levels of eight main ABC transporters in brain (cortex) and lateral choroid plexuses were investigated in chronically treated animals and compared to samples from acute experiments. Throughout this paper, the ABC transporter terminology will reflect the experiment described. For gene-based studies (RT-qPCR), genes will be listed with common protein names in brackets, e.g. abcc1 (MRP1), whereas for protein-based experimentation, the protein name will be listed with associated gene in brackets, e.g. MRP1 (abcc1). The full raw data are available as Underlying data (Habgood et al., 2019).\n\nAcute experiments: a single dose of drug mixed with a tracer amount of radiolabelled drug was injected (i.p.). Chronic experiments: twice daily injections (i.p.) over five days of “cold” drug with a tracer amount of radiolabelled drug included in the final dose. Numbers (n) are the same as in Table 1 and indicate samples for brain and plasma, with values in brackets indicating numbers for CSF samples when they differed from those for brain and plasma. Adults were non-pregnant females and littermates of both sexes were included in the E19 and P4 age groups of pups. SD=standard deviation.\n\nFor n numbers see Table 5. Adults were non-pregnant females, E19 and P4 pups of both sexes were littermates respectively.\n\nBars are group means with individual data points shown, * p<0.05, ** p<0.01. Note that for both acute and chronic treatment groups, the ratios are higher in brain and CSF in younger animals. At E19, digoxin was administered by i.p. injection to the mother. Individual fetuses were serially sampled starting at 30 minutes following maternal injection up to approximately 2.5 hours (see Figure 6 for times of sampling and maternal and fetal plasma digoxin levels). Adult and P4 animals were injected i.p. and samples taken at 30 minutes.\n\nBars are group means with individual data points shown. At E19, cimetidine was administered by i.p. injection to the mother. Individual fetuses were serially sampled starting at 30 minutes following maternal injection up to approximately 2.5 hours (see Figure 6 for times of sampling and maternal and fetal plasma cimetidine levels). Adult and P4 animals were injected i.p. and samples taken at 30 minutes. Note that for both acute and chronic treatment groups, the ratios are higher in brain and CSF in younger animals. There were no significant differences between acute and chronic groups at any of the three developmental ages investigated.\n\nBars are group means with individual data points shown, * p<0.05, p<0.01, * P<0.001. At E19, paracetamol was administered by i.p. injection to the mother. Individual fetuses were serially sampled starting at 30 minutes following maternal injection up to approximately 2.5 hours (see Figure 6 for times of sampling and maternal and fetal plasma paracetamol levels). Adult and P4 animals were injected i.p. and samples taken at 30 minutes. Note that for both acute and chronic treatment groups, the ratios were higher in brain and CSF in younger animals. At E19, following chronic doses, ratios for both brain and CSF were higher; in the adults they were lower.\n\nDigoxin. In acute digoxin experiments (Figure 1), the brain/plasma concentration ratio at E19 (47%) was much higher than the corresponding CSF/plasma ratio (12%). At P4, the brain/plasma (20%, p<0.05) and CSF/plasma (4%, p<0.05) ratios were both significantly lower than at E19. In the adults, the brain/plasma ratio declined further to 12% (p=0.06), while the CSF/plasma ratio remained at its already low level (4%). In chronic experiments (Figure 1), the brain/plasma and CSF/plasma ratios were not significantly different from the acute experiments at E19 (45% and 18%, respectively) or at P4 (18% and 3%, respectively). In contrast, in adults following chronic administration, brain/plasma concentration ratio was significantly lower compared to acute experiments (5%, compared to 12%, p<0.01) and the CSF/plasma ratio also decreased significantly (3%, p<0.05).\n\nCimetidine. In acute cimetidine experiments (Figure 2), the brain/plasma and CSF/plasma concentration ratios were highest at E19 (56% and 54%, respectively). At P4, the brain/plasma (12%, p<0.001) and CSF/plasma (8%, p<0.001) ratios were significantly lower than at E19. In adults, the brain/plasma (13%) and CSF/plasma (12%) ratios were not statistically different from P4. There were no significant differences observed between the acute and chronic experiments for brain/plasma or CSF/plasma concentration ratios at any age.\n\nParacetamol. In acute (single dose) experiments (Figure 3), the brain/plasma and CSF/plasma ratios were highest at E19 (66% and 60%, respectively). At P4, ratios were lower (60% and 49%, respectively), with the CSF/plasma concentration ratio significantly different from that of E19 (p<0.05). In adults, both ratios were substantially lower than at P4 and E19, at 30% and 29%, respectively (p<0.001 for all comparisons). In chronic experiments (Figure 3) at E19, the ratios were substantially higher than in the acute experiments for both brain (104%, p<0.01) and CSF (85%, p<0.05). At P4, ratios in chronically treated animals were lower than in acute experiments for brain (51%) and CSF (43%, p<0.05), with the CSF/plasma concentration ratio significantly different from acute data. In adults, both ratios were substantially lower following chronic treatment compared with acute experiments (brain 21%, p<0.05; CSF 18%, p<0.001).\n\nTranscripts of all eight ABC transporters investigated were detected in brain cortex and choroid plexuses at all ages studied and in placental tissue at E19 (Table 7). In adult brain cortex, chronic digoxin exposure resulted in a significant up-regulation of abcb1a (PGP) expression (1.21 fold, p<0.05). In contrast, at E19, no significant up-regulation was observed in brain cortex. Instead, chronic digoxin treatment resulted in a significant down-regulation of abcb1b (PGP) expression (0.63 fold, p<0.05). At P4, the response to chronic drug exposure in the brain was more variable. For all three drugs there was a down-regulation of abcc2 (MRP2; Table 7), as well as an up-regulation of abcg2 (BCRP) following chronic paracetamol (1.52 fold, p<0.05) and abcc5 (MRP5) following chronic digoxin (1.15 fold, p<0.05) and chronic cimetidine (1.24 fold, p<0.01) exposure. In the choroid plexus samples at all three ages, regulation in response to chronic drug exposure appeared to be variable. Regulation was different within each age for the three drugs, as well as for each drug between ages (Table 7). Possible correlations between ABC transporters expression and drug entry into the brain will be considered in the Discussion.\n\nThe results are fold change differences between the chronic and acute treatments. They confirm the expression of MRP1-5 (abcc1-5), BCRP (abcg2) and P-glycoprotein/PGP (abcb1a, abcb1b) in these rat tissues. Statistically significant differences in transcript numbers in the chronic treatment group (p<0.05) are indicated by ↑ where there was up-regulated expression of the target and by ↓ where there was down-regulated expression. * indicates a large fold change increase that did not reach statistical significance due to extremely low expression in all acute and chronically treated animals except for two animals that expressed the transporter at a high level. Note the near total absence of changes in expression in brain and choroid plexus at E19 compared to the large number of changes at P4.\n\nAccess of any molecule into the developing brain and CSF is determined by its physical, chemical and physiological properties and the nature of the barrier interfaces at the time (see Introduction). Therefore, a comparison was made between the drug permeability results in this study with measurements of “passive” markers of similar molecular size (L-glucose, sucrose and glycerol). These markers are thought to not bind to any influx or ABC efflux transporters and have varying lipid solubility (log Doctonol). The brain/plasma (Figure 4) and CSF/plasma (Figure 5) ratios at 30 minutes after i.p. injection were determined for radiolabelled L-glucose, sucrose, and glycerol at E19, P4 and in adults. The ratios are plotted against log Doctonol for each marker and for each drug, with higher lipid solubility, known to have the potential to result in increased barrier permeability (see Discussion). The brain/plasma and CSF/plasma ratios of the hydrophilic markers sucrose and L-glucose were very low at all ages, whereas the ratio for the more lipophilic glycerol was about 50% at E19 (46.8% brain and CSF) and approached 100% at P4 (82.2% brain and 95% CSF) and in adults (85.5% brain and 103.9% CSF). In separate experimentation, 60-minute glycerol concentrations ratios reached approximately 80% in the brain and nearly 100% in the CSF at E19 (see Discussion), making values very similar to those of P4 and adult. The comparison of the brain/plasma (Figure 4) and CSF/plasma (Figure 5) ratios for the “passive” markers and the drugs at each age suggests drug exclusion, most likely by efflux transport mechanisms. By being excluded, we mean that the brain or CSF to plasma ratios are much lower than would be expected from their LogDoctonol position in the figure. However, this relation only applies directly to a comparison between CSF and plasma, as both compartments are water-based. The results for brain to plasma ratios can only be an indication, as both compartments are different in terms of their cellular composition; brain distribution space is a combination of intracellular and extracellular compartments.\n\n(A) digoxin, (B) cimetidine and (C) paracetamol in E19 fetal (green triangles), P4 postnatal (blue squares) and adult (red circles) rats plotted against their lipid solubility (LogDOctanol partition coefficient) and compared with the “passive” permeability markers sucrose, L-glucose and glycerol. Filled symbols indicate acute experiments (30 minutes after IP injection) and open symbols indicate chronic experiments (after twice daily IP injections over five days). See Table 1 and Table 6 for full data and n numbers. Ratios less than 100% indicate restricted entry of the drug or marker into brain. Digoxin, despite being the most lipid soluble of the drugs was the most restricted at all ages in both the acute and chronic treatments. Cimetidine was similarly restricted at P4 and in adults, but less so at E19. Paracetamol was the least restricted of the drugs. Also note that following chronic treatment, paracetamol entry into brain decreased in adults, but was increased at E19 (direction of significant changes indicated by arrows). Chronic treatment also reduced the entry of digoxin into adult brain adults, but not at P4 and E19.\n\nAs can be seen in the brain (Figure 4) and in the CSF (Figure 5), despite having the highest lipid solubility, digoxin transfer at P4 and in adults was at a level similar to the hydrophilic sucrose and L-glucose, suggesting active barrier exclusion of digoxin. If the transfer of digoxin was unobstructed, based on its logDoctonol value, it would be expected to reach much higher ratios at or above those for glycerol. For cimetidine, the exclusion was similar to that of digoxin at P4 and adult. In contrast, digoxin and cimetidine transfer at E19 was similar to that of glycerol. As their lipid solubility is much higher, it would be expected that, if passively transferring, their transfer ratios would be higher than glycerol; therefore, it seems likely that the degree of exclusion of digoxin and cimetidine at E19 is less than what was observed at P4 and adult. Paracetamol was generally less excluded than digoxin and cimetidine at all ages. The most striking difference was at E19 following chronic treatment, when it seems that this drug was not excluded at all (i.e. reached 100% distribution ratio between brain/CSF and plasma). In contrast to E19, the chronic treatment regime in the adult resulted in reduced entry of this drug (Figure 4). Possible reasons for this age-related difference in paracetamol ratios are considered in the Discussion.\n\n(A) digoxin, (B) cimetidine and (C) paracetamol in E19 fetal (green triangles), P4 postnatal (blue squares) and adult (red circles) rats plotted against their lipid solubility (LogDOctanol partition coefficient) and compared with the “passive” permeability markers sucrose, L-glucose and glycerol. Filled symbols indicate acute experiments (30 minutes after IP injection) and open symbols indicate chronic experiments (after twice daily IP injections over five days). See Table 1 and Table 6 for full data and n numbers. Ratios less than 100% indicate restricted entry of the drug or marker into CSF. Digoxin, despite being the most lipid soluble of the drugs, was the most restricted at all ages in both the acute and chronic treatments. Cimetidine was similarly restricted at P4 and in adults, but less so at E19. Paracetamol was the least restricted of the drugs. Also note that following chronic treatment, paracetamol entry into CSF decreased in adults, but was increased at E19 (direction of significant changes indicated by arrows). Chronic treatment did not significantly affect the entry of cimetidine into CSF at any of the three ages investigated.\n\nIn summary, it is clear that the brain/plasma and CSF/plasma ratios were much higher in the E19 fetuses, both for acute and chronic experiments. The highest ratios obtained were those for paracetamol. The concentration ratios decreased substantially for all three drugs by P4, although less so for paracetamol. For this drug there was a further decrease in adult brain and CSF ratios (Figure 1–Figure 3). This trend appeared to be specific to the lipid soluble, efflux substrate drugs and not due to general barrier permeability changes as glycerol did not follow the same trend. Entry of all three drugs into both the brain and the CSF was lower (except at E19 in chronic paracetamol experiments, see Discussion) than could be predicted from their LogDoctanol if their transfer was entirely unrestricted (Figure 4 and Figure 5).\n\nAn estimate of the placental transfer of drug molecules at E19 was obtained by comparing the maternal and fetal plasma levels of radiolabelled drugs as illustrated in Figure 6. Maternal plasma was sampled from an arterial cannula periodically throughout the experiment (see Methods). The fetal values are from individual fetuses at the termination of their exposure to radiolabelled drug. A 2.5h post-injection cut off was established to ensure these pups were collected whilst there was good placental circulation (see Methods). As can be seen in Figure 6, the maternal plasma level of all three drugs declined throughout the experimental period but remained consistently higher than fetal plasma drug levels. Fetal plasma values followed different patterns for all three drugs. The fetal plasma paracetamol level declined with time, but at a rate that was slower compared to the maternal plasma (Figure 6). Digoxin and cimetidine fetal plasma levels were relatively stable throughout the experiment. In order to obtain the rate of placental transfer for each drug, individual fetal plasma values were time matched to the maternal plasma samples. An average of the two nearest maternal plasma values was used where there was not a direct time match. These ratios are presented in Table 8. For all three drugs, the placenta was restricting the transfer by about 60% (paracetamol/digoxin) to 70% (cimetidine). This is indicated by the ratios of fetal to maternal plasma levels; 42% (paracetamol), 37% (digoxin) and 30% (cimetidine) (Table 8). No significant difference was found between the transfer of the drugs across the placenta between acute and chronic administration. There were, however, some changes in placental ABC transporter expression in response to chronic treatment (Table 7 and below).\n\nNote that the maternal plasma levels for both acute and chronic experiments declined progressively for all three drugs throughout the 2–2.5 hours experimental period, but the fetal plasma levels were stable for digoxin and cimetidine during the same period. The paracetamol levels in fetal plasma declined during this period. For digoxin and cimetidine, the levels of radiolabelled drugs in acute and chronic experiments were similar, but for paracetamol the levels in maternal and fetal plasma were much higher with chronic treatment. The much lower levels for each drug in fetal plasma indicates a substantial restriction of drug transfer across the placenta. Lines fitted by Least Squares Linear Regression, curve fitted by Least Squares Exponential Decay (one phase).\n\nData shown are mean ± SD fetal/maternal plasma concentration ratios (%) for the radiolabelled tracers in acute (drugs and permeability markers) and chronic (drugs only) experiments. The range of the ratios obtained are shown in brackets.\n\nDespite no changes being observed in transfer of paracetamol, digoxin or cimetidine across the placenta following chronic treatment, some differences were observed in ABC-transporter expression (Table 7). In placentas from the paracetamol treated dams, abcc1 (MRP1) and abcg2 (BCRP) were up-regulated (1.69 fold, p<0.05 and 2.0 fold, p<0.01 respectively). In contrast, abcc1 (MRP1) was down-regulated following chronic treatment with digoxin or cimetidine (0.77 fold, p<0.05 and 0.87 fold, p<0.05 respectively) and abcb1b (PGP) was down-regulated only following digoxin treatment (0.47 fold, p<0.05).\n\nIn the E19 animals, passive markers (L-glucose and glycerol) indicated the extent to which placenta was able to restrict passive entry of small water-soluble markers from maternal blood and into the fetal circulation (Figure 7) during the experiments conducted. The levels of the hydrophilic marker L-glucose in the fetal plasma was low compared to the maternal plasma (16.7% ± 6.5%) during the 30 minute experimental period, whereas the more lipophilic glycerol reached 100% of the maternal plasma, indicating unrestricted transfer (Table 9).\n\n[3H]-labelled “passive” permeability markers L-glucose and glycerol (both open circles) are shown; along with [3H]-labelled drugs digoxin (blue triangles), cimetidine (purple diamond) and paracetamol (pink hexagon). Filled symbols indicate acute experiments and open symbols indicate chronic experiments. Values are means, see Table 8 for full data and Table 1 and Table 2 for n numbers. Note that all three drugs showed a degree of restriction (values below the trend of passive permeability) provided by the placental barrier. There was no difference between acute and chronic treatment groups for any of the drugs tested.\n\nIn Figure 7, for comparative purposes, the values for the placental transfer of the three drugs are also included. Because of their high lipophilicity, it could be predicted that if they were not being actively excluded, their maternal to fetal transfer ratios would have also been close to 100%. As illustrated in Figure 7, paracetamol, digoxin and cimetidine ratios were all much lower, indicating that the placenta was able to partly impede their passage, both in acute and chronic treatment groups. Potential effects of placental exclusion on the levels of drugs reaching the fetal brains from the maternal circulation are considered in the Discussion.\n\n\nDiscussion\n\nIn this study we aimed to determine the level of entry into the brain and CSF of three drugs at different ages. We also assessed the age-dependent functional capacity of brain barriers to prevent or limit the entry of potentially harmful drugs and how this may alter with a longer treatment regime. There is only limited information available about the expression and cellular distribution of key barrier defences such as ABC transporters in the developing brain, either in the human embryo/fetus (Møllgård et al., 2017) or in developing rodents (Ek et al., 2010). Whether these transporters are functionally active early in development or how chronic drug exposure might affect their functionality has not yet been investigated. In this paper, drug entry has been studied at fetal day 19 (E19), postnatal day four (P4) and in adult rats using clinically relevant doses of paracetamol, cimetidine and digoxin, either as a single injection or as multiple administration over five days. Well-established methods were employed to measure radiolabelled drugs/markers in the brain, blood and CSF of fetal and neonatal animals (Dziegielewska et al., 1979; Dziegielewska et al., 1982; Habgood et al., 1993; Johansson et al., 2006; Stolp et al., 2005). This has been combined with RT-qPCR to determine any changes in gene expression level of the ABC transporters that have been shown to be functionally important in the adult brain (Roberts et al., 2008; Saidijam et al., 2018) in response to chronic drug treatment.\n\nThe rat has been chosen because there is considerable knowledge about brain development in this species compared with humans (Clancy et al., 2001). In particular, the stage of brain development of the cerebral cortex (the main region studied) in early postnatal rats is similar to that of human fetuses at 22–24-week gestation (Clancy et al., 2001). This is important because just past the mid-gestation is the earliest stage of viability for pre-term birth (Fischer et al., 2009; Stoll et al., 2010), making the rat model noteworthy for investigating a time where the developing brain may be particularly vulnerable because of the loss of placental protection.\n\nPermeability of the drugs into the brain and CSF was compared to that of similarly small-sized molecules, but with different diffusion coefficients and that are not actively transported, i.e. they enter the brain and CSF by “passive” diffusion only. This provides a comparison that is useful in evaluating the level of restriction of the drug’s penetration at different ages (the functionality of the relevant ABC transporter). As indicated above, the level of drug or marker attained in the brain in these experiments is a measure of “apparent permeability” rather than absolute permeability because of the influence of the turnover of CSF, which is much less in the developing brain (Saunders et al., 1992). Comparison of drug entry and passive marker entry allows interpretation to take account of possible effects of CSF turnover.\n\nThere are very few published studies of drug permeability of the developing brain when they were administered to pregnant rats or mice. Of these, fluoxetine and venlafaxine administered via drinking water between E0 and E10 could not be detected in the brains of the embryos (Kaushik et al., 2016). All of the other drugs (digoxin, saquinavir, paclitaxel, cimetidine, apipaxaban, mitoxaurone, talinolol, carbamazepine, genisten, genisten, daidzein and coumestrol), which were administered i.v. or orally between E15 and E21, were detected in the fetal brain (Cygalova et al., 2008; Enokizono et al., 2007; Kaushik et al., 2016; Petropolous et al., 2010; Petropolous et al., 2011; Saljé et al., 2012; Smit et al., 1999). However, neither fetal blood nor CSF were sampled and the brain level of the drug (usually radiolabelled) was related to maternal blood or to other fetal tissues. This makes it difficult to assess the contribution of the different barrier interfaces (placenta, blood-brain and blood-CSF barriers) that may be contributing to limiting entry of the drugs into the developing brain. The present study provides information on this, as both the maternal and fetal blood and CSF were sampled in addition to the brain tissue.\n\nThere are also very few investigations into the regulatory capacity of the developing brain in response to chronic drug exposure. Previous studies have shown that, following chronic drug challenge, blood-brain interfaces can increase their levels of protection (ABC transporters), which results in greater efflux capacity of compounds (Cui et al., 2009; Hoque et al., 2015). Our recent study suggested that adult animals may have a greater capacity to upregulate ABC transporters at blood-brain interfaces following chronic xenobiotic exposure than early in development (Koehn et al., 2019).\n\nDigoxin has generally been found to be a substrate for PGP (abcb1). Increased brain/plasma ratios for digoxin have been reported in wild type mice when co-administered with a PGP inhibitor and also when administered to mdr1a/1b (PGP) knockout mice, indicating that this transporter is of particular importance in limiting entry of digoxin into the brain (Mayer et al., 1997). Petropoulos et al. (2010) suggested that levels of radiolabelled digoxin (compared to the rest of the fetus) were higher earlier in gestation when fetal levels of abcb1a (PGP) were lower. Smit et al. (1999) reported a fetal brain to maternal plasma ratio of 25% (4hrs) and 100% (24hrs) after i.v. injection of radiolabelled digoxin in pregnant mice. These would have been overestimates of the brain level, as the maternal blood level would fall progressively following injection. This type of experiment requires a dosage regime that maintains an approximately constant blood level of marker (Dziegielewska et al., 1979) for a realistic estimate of the brain level to be obtained. In the present study, digoxin transferred into the adult and P4 brain (10–20%) and CSF (4%) at much lower levels compared to E19 (brain 47%, CSF 12%). The much lower entry into CSF suggests that the efflux mechanisms are more effective in the choroid plexuses than in the brain itself, even as early as E19. Following chronic treatment, transfer into the adult brain decreased (12% to 5%) in a manner that correlated with increased expression of abcb1a in the cerebral cortex. This result correlates with the above finding that digoxin is likely to be a PGP (abcb1) substrate. As this up-regulation (and functional transfer decrease) only occurred in adults and not at P4 or E19, the results are consistent with those described by Koehn et al. (2019). In both studies, chronic exposure to a PGP inducer up-regulated abcb1a expression in the adult brain but not earlier in development, suggesting that for a range of molecular inducers, regulation of blood-brain barrier defences may be age-dependent.\n\nCimetidine has been typically linked to BCRP (abcg2) as its efflux mechanism. Experiments in dually perfused rat placenta suggested BCRP (abcg2), but not PGP (abcb1a/b), involvement in cimetidine efflux (Staud et al., 2006) and changes in abcg2 (BCRP) expression in the rat brain have also been linked to differential transfer of cimetidine (Liu et al., 2007). Cygalova et al. (2008) found that the level of radiolabelled cimetidine in the fetal brain at E18 or E21 in pregnant rats one hour after i.v. infusion was less in the older fetuses but abcg2 (BCRP) mRNA expression was significantly less at E21 compared to E18, so presumably some other protective mechanisms may have been involved. In the present study, in acute experiments, cimetidine entry into the adult brain was similar to that of digoxin (13%) but entered the CSF to a higher ratio (12% compared to digoxin 4%). The higher degree of cimetidine transfer into the CSF compared to digoxin could be due to less ABC transporter efflux capacity at the choroid plexus or from its lower lipid solubility, allowing greater partition into the aqueous CSF fluid. There do not appear to be any cimetidine-induced changes in efflux capacity at different ages, as no differences in transfer were detected in chronically compared to acutely treated animals. Consistent with this finding are the RT-qPCR results, showing only limited changes in efflux transporter expression following chronic cimetidine treatment (Table 7). Most changes in ABC transporter expression in response to cimetidine occurred in the brain of P4 animals. In fact, this early postnatal period of rat brain development is characterised by increased expression of some efflux transporters in response to each of the three drugs and could indicate a maturation process that is not yet fully established (see Table 7).\n\nSo far, paracetamol has not been clearly linked to any specific efflux transport mechanism, although it is known to be metabolised via glucaronidation, sulfation and (via an intermediate) glutathionation, making it likely to interact with BCRP (abcg2) and the family of MRPs (abcc; Mazaleuskaya et al., 2015). In adult rats, Courade et al. (2001) found brain/plasma ratios for 3H-paracetamol of around 40% at 45 minutes after i.v. administration in different brain regions; this is similar to our value after 30 minutes (Figure 3). In the present study, paracetamol’s brain/plasma and CSF/plasma ratios were considerably higher than digoxin and cimetidine at E19 (66% and 60%, respectively), P4 (~60% and ~50%, respectively) and in the adult (~30% for both). This suggests that the mechanisms preventing paracetamol entry may be less effective than those targeting digoxin and cimetidine. As paracetamol has a lower lipid solubility than digoxin (Figure 4 and Figure 5) and therefore a predicted lower barrier permeability (Levin, 1980), this result is even more pronounced. Most interesting were the results obtained following chronic paracetamol treatment. In the adult and P4 chronically treated animals, the brain/plasma and CSF/plasma ratios decreased by approximately 10%. However, no clear regulatory mechanism could be established from RT-qPCR results, as only abcg2 (BCRP) up-regulated at the P4 brain and abcb1b (PGP) up-regulated in the adult choroid plexus (Table 7). It is therefore possible that the up-regulation occurred due to other aspects of efflux mechanisms such as the metabolising enzymes required to conjugate glucuronic acid, sulphate or glutathione groups onto paracetamol for efflux by the appropriate transporters. In fetal animals the opposite effect was observed, with an increase in both ratios to around 100%, suggesting that there was no restriction on paracetamol entry. This might indicate that brain and CSF-barrier efflux capacity was exceeded, resulting in accumulation of paracetamol in the E19 fetuses. Experiments to better understand the mechanisms regulating the entry of paracetamol into the developing brain are in progress.\n\nThe level of a drug or other marker in brain following administration depends on the duration of the experiment, its diffusion coefficient (D), its lipid solubility (Log Doctonol), the turnover of CSF (sink effect) and the effectiveness of a specific transport mechanism if present; the latter could be inward, as in the case of amino acids and a small number of drugs, but generally outward for drugs that are substrates for ABC efflux transporters.\n\nAs has been shown in multiple studies, one major factor in the degree of the transfer is the lipid solubility of the compound. For compounds that pass barriers “passively” without active efflux, as lipid solubility increases, the transfer across the barrier interfaces should increase (Garber et al., 2005; Levin, 1980). The graphical data in Figure 5 and Figure 6 display each compound’s brain/plasma and CSF/plasma ratios against their lipid solubilities. For E19, P4 and adults, brain to plasma concentration ratios for passive markers (sucrose, L-glucose, glycerol) increased as lipid solubility increased.\n\nThe comparison between the brain entry of the three drugs with that of markers that do not bind to efflux transporters provides valuable insight, as most factors contributing to barrier permeability should be the same, except for active transport. The increased drug transfer earlier in development (described above) appears not to be a general property of the barrier for all molecules as glycerol did not follow this trend. This, combined with the distance of the three lipid soluble drugs below the line of passive transfer (Figure 5 and Figure 6), indicate changes in the functional capacity of efflux transporters at the barriers over development.\n\nIn our experiments we have been able to obtain an estimate of the placental contribution to the overall protection of the fetus from drugs administered to the pregnant mother by comparing directly the drug levels in fetal and maternal plasma. The fetal/maternal plasma ratios varied between ~30% for cimetidine treatment to ~40% for paracetamol and digoxin (Figure 6 and Figure 7; Table 8), indicating a substantial protective barrier for these drugs provided by the placenta in late gestation in the rat. The placenta did not, however, completely prevent molecular transfer. RT-qPCR estimates of ABC transporter expression in the rat placenta at E19 confirmed the presence of abcb1a, abcb1b, abcg2 and abcc1-5 in varying quantities, as has been shown in several published studies (Kalabis et al., 2007; Leazer et al., 2003; Novotna et al., 2004). There were only a few small changes in expression in these transporters following chronic treatment with the drugs (Table 7). Chronic paracetamol resulted in significant up-regulation of abcc1 (MRP1) and abcg2 (BCRP), while digoxin and cimetidine caused a down-regulation of abcc1. However, these changes were not reflected in changes in placental transfer. This result may be of clinical importance as it demonstrated differences in the regulation of ABC transporters expression that are tissue specific: up-regulation with associated decreases in functional transfer into the brain, but not in the placenta.\n\nTaking into account the level of drug exclusion provided by the placenta for all three drugs, the transfer from maternal blood to fetal brain in acute experiments was at a relatively similar concentration ratio as that from maternal blood to the maternal brain. Digoxin transfer into maternal brain was 12%, whereas the transfer from maternal blood to fetal brain was 17%. Similar results were seen for paracetamol (30% maternal brain, 28% fetal brain) and cimetidine (13% maternal brain, 17% fetal brain). In contrast to the acute experiments, different patterns of transfer were observed into the maternal and fetal brains following chronic treatment. Cimetidine, which did not up-regulate ABC transporter expression or decrease its transfer with chronic treatment, had the same ratio of transfer into both the maternal and fetal brains. Digoxin, however, decreased its entry into the maternal brain during chronic treatment, while the entry across the placenta and into the fetal brain remained the same. Paracetamol showed decreased entry into the maternal brain, while the amount entering fetal brain increased. For acute paracetamol treatment, transfer into brain (brain/maternal blood) was 30% for mother and 28% for the fetus, but after five days of twice-daily paracetamol, the transfer of a dose on the 5th day was 21% for the mother and 45% for the fetus. Thus, in chronic treatment, paracetamol may have become less effective for the mother, but potentially more harmful for the fetus as more paracetamol reached its brain. If the same situation applies in patients, it would suggest that, where possible, duration of treatment should be as short as possible.\n\nThis study has been restricted to two treatment conditions; namely, a single acute dose and twice daily chronic treatment over a period of five days. The latter corresponds to about one third of gestation. It is possible that with an even longer period of treatment, greater effects on ABC transporter expression correlated with greater restriction of drug entry in fetuses, neonates and adults would have been observed. We studied only one dose concentration of each drug but this was chosen to be in the clinical range. It is possible that with smaller doses, a lower rate of entry would be obtained. This could be the case particularly in our paracetamol experiments at E19; it seems likely that the greater entry of paracetamol may have been due to the capacity of the efflux transporters for this drug having been exceeded.\n\nWe have estimated drug entry using only radiolabelled drugs, rather than direct measurement of unlabelled drugs, as the amounts of fetal/postnatal fluids would have been too small for other methods such as high-performance liquid chromatography (HPLC). However, the similarity of our results in adult rats for paracetamol and those of Courade et al. (2001), who measured paracetamol using HPLC, suggests that the isotopically labelled drugs gave a reliable index of permeability in our experiments. There are always legitimate concerns about the extent to which results from animal studies can be extrapolated to humans. In terms of brain development, we chose P4 because of the general similarity of brain development in the rat to very preterm human infants (Clancy et al., 2001). There are differences in the detail of the structure of the placentas in humans and rats, but they are both haemochorial and are much more comparable for studies such as those in this paper than, for example, the much “tighter” epitheliochorial multicotyledonary placenta of the ewe (Studert et al., 2011) that has been used for a lot of developmental studies.\n\nOur results show that at the doses used, all three drugs entered the brain at all three ages studied. The entry of all of the drugs was highest in the youngest animals (E19). This is probably a combination of the negligible turnover of CSF at this age (Saunders, 1992), allowing greater accumulation and possibly lesser activity of the relevant active efflux transporters. The entry into brain was also appreciably higher for paracetamol than the other two drugs at all three ages. The finding of entry of digoxin and cimetidine into the brain, particularly at E19, suggests that these drugs should be studied for possible long-term effects on brain development and behaviour of offspring. However, a much greater entry of paracetamol suggests that experiments to test for such effects in the offspring would be particularly appropriate, since this is the drug most commonly taken by pregnant women (Werler et al., 2005). A more complex matter that requires investigation is the possibility of multiple drug administration having untoward effects due, for example, to interactions with common ABC transporter efflux mechanisms.\n\nThis study provides an experimental basis for future examination of other drugs administered in pregnancy. If the extent of drug entry into the developing brain and its regional distribution can be established, this would create a foundation for studying potential deleterious effects of drugs on brain development and possible related changes in postnatal behaviour. In addition, paracetamol and digoxin results highlight the potential differences in how adults, neonates and fetuses respond to chronic drug exposure. Adults appear more capable of responding to drug challenge by regulating ABC transporter activity, thereby allowing less drug transfer into the CSF and brain.\n\n\nData availability\n\nFigshare: Determinants of drug entry into the developing brain: raw data files. https://doi.org/10.26188/5d3e5539dca74 (Habgood et al., 2019)\n\nThis project contains the following underlying data:\n\n- Koehn et al., 2019 qPCR Raw Data.xlsx\n\n- Koehn et al., 2019 Sucrose Raw Data.xlsx\n\n- Koehn et al., 2019 L-Glucose Raw Data.xlsx\n\n- Koehn et al., 2019 Glycerol Raw Data.xlsx\n\n- Koehn et al., 2019 Paracetamol Raw Data.xlsx\n\n- Koehn et al., 2019 Digoxin Raw Data.xlsx\n\n- Koehn et al., 2019 Cimetidine Raw Data.xlsx\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nThis work was supported by the CASS Foundation, Victoria, Australia [7981].\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nAdedoyin A, Aarons L, Houston JB: Dose-dependent pharmacokinetics of cimetidine in the rat. Xenobiotica. 1987; 17(5): 595–604. PubMed Abstract \| Publisher Full Text\n\nAustralian Medicines Handbook Pty Ltd. Adelaide SA, Australia. 2019. Reference Source\n\nBito LZ, Bradbury MW, Davson H: Factors affecting the distribution of iodide and bromide in the central nervous system. J Physiol. 1966; 185(2): 323–354. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBriggs GG, Freeman RK, Towers CV, et al.: Drugs in Pregnancy and Lactation: A Reference Guide to Fetal and Neonatal Risk. 11th edn. Lippincott Williams & Wilkins (LWW), Philadelphia. 2017. Reference Source\n\nClancy B, Darlington RB, Finlay BL: Translating developmental time across mammalian species. Neuroscience. 2001; 105(1): 7–17. PubMed Abstract \| Publisher Full Text\n\nCourade JP, Besse D, Delchambre C, et al.: Acetaminophen distribution in the rat central nervous system. Life Sci. 2001; 69(12): 1455–1464. PubMed Abstract \| Publisher Full Text\n\nCui YJ, Cheng X, Weaver YM, et al.: Tissue distribution, gender-divergent expression, ontogeny, and chemical induction of multidrug resistance transporter genes (Mdr1a, Mdr1b, Mdr2) in mice. Drug Metab Dispos. 2009; 37(1): 203–210. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCygalova L, Ceckova M, Pavek P, et al.: Role of breast cancer resistance protein (Bcrp/Abcg2) in fetal protection during gestation in rat. Toxicol Lett. 2008; 178(3): 176–180. PubMed Abstract \| Publisher Full Text\n\nDavson H: Physiology of the Cerebrospinal Fluid. Churchill Livingston, Edinburgh. 1967. Reference Source\n\nDavson H, Segal MB: Physiology of the CSF and blood-brain barriers. CRC press, Boca Raton. 1996. Reference Source\n\nDziegielewska KM, Evans CA, Lai PC, et al.: Proteins in cerebrospinal fluid and plasma of fetal rats during development. Dev Biol. 1981; 83(1): 193–200. PubMed Abstract \| Publisher Full Text\n\nDziegielewska KM, Evans CA, Malinowska DH, et al.: Studies of the development of brain barrier systems to lipid insoluble molecules in fetal sheep. J Physiol. 1979; 292: 207–231. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDziegielewska KM, Saunders NR: Transferrin in fetal sheep cerebrospinal fluid and plasma during gestation. Comp Biochem Physiol A Comp Physiol. 1982; 73(2): 327–329. PubMed Abstract \| Publisher Full Text\n\nEk CJ, Wong A, Liddelow SA, et al.: Efflux mechanisms at the developing brain barriers: ABC-transporters in the fetal and postnatal rat. Toxicol Lett. 2010; 197(1): 51–59. PubMed Abstract \| Publisher Full Text\n\nEnokizono J, Kusuhara H, Sugiyama Y: Effect of breast cancer resistance protein (Bcrp/Abcg2) on the disposition of phytoestrogens. Mol Pharmacol. 2007; 72(4): 967–975. PubMed Abstract \| Publisher Full Text\n\nFischer N, Steurer MA, Adams M, et al.: Survival rates of extremely preterm infants (gestational age <26 weeks) in Switzerland: impact of the Swiss guidelines for the care of infants born at the limit of viability. Arch Dis Child Fetal Neonatal Ed. 2009; 94(6): F407–F413. PubMed Abstract \| Publisher Full Text\n\nHabgood MD, Knott GW, Dziegielewska KM, et al.: The nature of the decrease in blood-cerebrospinal fluid barrier exchange during postnatal brain development in the rat. J Physiol. 1993; 468: 73–83. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHabgood M, Koehn L, Huang Y, et al.: Determinants of drug entry into the developing brain: raw data files (Version 1). University of Melbourne. 2019. http://www.doi.org/10.26188/5d3e5539dca74\n\nHarrison LI, Gibaldi M: Pharmacokinetics of digoxin in the rat. Drug Metab Dispos. 1976; 4(1): 88–93. PubMed Abstract\n\nHoque MT, Shah A, More V, et al.: In vivo and ex vivo regulation of breast cancer resistant protein (Bcrp) by peroxisome proliferator-activated receptor alpha (Pparα) at the blood-brain barrier. J Neurochem. 2015; 135(6): 1113–1122. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJohansson PA, Dziegielewska KM, Ek CJ, et al.: Blood-CSF barrier function in the rat embryo. Eur J Neurosci. 2006; 24(1): 65–76. PubMed Abstract \| Publisher Full Text\n\nJohansson PA, Dziegielewska KM, Liddelow SA, et al.: The blood-CSF barrier explained: when development is not immaturity. BioEssays. 2008; 30(3): 237–248. PubMed Abstract \| Publisher Full Text\n\nKalabis GM, Petropoulos S, Gibb W, et al.: Breast cancer resistance protein (Bcrp1/Abcg2) in mouse placenta and yolk sac: ontogeny and its regulation by progesterone. Placenta. 2007; 28(10): 1073–1081. PubMed Abstract \| Publisher Full Text\n\nKaushik G, Huber DP, Aho K, et al.: Maternal exposure to carbamazepine at environmental concentrations can cross intestinal and placental barriers. Biochem Biophys Res Commun. 2016; 474(2): 291–295. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKodaira H, Kusuhara H, Fujita T, et al.: Quantitative evaluation of the impact of active efflux by p-glycoprotein and breast cancer resistance protein at the blood-brain barrier on the predictability of the unbound concentrations of drugs in the brain using cerebrospinal fluid concentration as a surrogate. J Pharmacol Exp Ther. 2011; 339(3): 935–944. PubMed Abstract \| Publisher Full Text\n\nKoehn LM, Dziegielewska KM, Møllgård K, et al.: Developmental differences in the expression of ABC transporters at rat brain barrier interfaces following chronic exposure to diallyl sulfide. Sci Rep. 2019; 9(1): 5998. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeazer TM, Klaassen CD: The presence of xenobiotic transporters in rat placenta. Drug Metab Dispos. 2003; 31(2): 153–167. PubMed Abstract \| Publisher Full Text\n\nLevin VA: Relationship of octanol/water partition coefficient and molecular weight to rat brain capillary permeability. J Med Chem. 1980; 23(6): 682–684. PubMed Abstract \| Publisher Full Text\n\nLin JH, Levy G: Effect of pregnancy on the pharmacokinetics of acetaminophen in rats. J Pharmacol Exp Ther. 1983; 225(3): 653–659. PubMed Abstract\n\nLiu X, Cheong J, Ding X, et al.: Use of cassette dosing approach to examine the effects of P-glycoprotein on the brain and cerebrospinal fluid concentrations in wild-type and P-glycoprotein knockout rats. Drug Metab Dispos. 2014; 42(4): 482–491. PubMed Abstract \| Publisher Full Text\n\nLiu YC, Liu HY, Yang HW, et al.: Impaired expression and function of breast cancer resistance protein (Bcrp) in brain cortex of streptozocin-induced diabetic rats. Biochem Pharmacol. 2007; 74(12): 1766–72. PubMed Abstract \| Publisher Full Text\n\nLyerly AD, Little OM, Faden R: The second wave: Toward responsible inclusion of pregnant women in research. Int J Fem Approaches Bioeth. 2008; 1(2): 5–22. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nManautou JE, de Waart DR, Kunne C, et al.: Altered disposition of acetaminophen in mice with a disruption of the Mrp3 gene. Hepatology. 2005; 42(5): 1091–1098. PubMed Abstract \| Publisher Full Text\n\nMandal A, Agrahari V, Khurana V, et al.: Transporter effects on cell permeability in drug delivery. Expert Opin Drug Deliv. 2017; 14(3): 385–401. PubMed Abstract \| Publisher Full Text\n\nMayer U, Wagenaar E, Beijnen JH, et al.: Substantial excretion of digoxin via the intestinal mucosa and prevention of long-term digoxin accumulation in the brain by the mdr 1a P-glycoprotein. Br J Pharmacol. 1996; 119(5): 1038–1044. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMayer U, Wagenaar E, Dorobek B, et al.: Full blockade of intestinal P-glycoprotein and extensive inhibition of blood-brain barrier P-glycoprotein by oral treatment of mice with PSC833. J Clin Invest. 1997; 100(10): 2430–2436. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMazaleuskaya LL, Sangkuhl K, Thorn CF, et al.: PharmGKB summary: pathways of acetaminophen metabolism at the therapeutic versus toxic doses. Pharmacogenet Genomics. 2015; 25(8): 416–426. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMøllgård K, Dziegielewska KM, Holst CB, et al.: Brain barriers and functional interfaces with sequential appearance of ABC efflux transporters during human development. Sci Rep. 2017; 7(1): 11603. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStaud F, Vackova Z, Pospechova K, et al.: Expression and transport activity of breast cancer resistance protein (Bcrp/Abcg2) in dually perfused rat placenta and HRP-1 cell line. J Pharmacol Exp Ther. 2006; 319(1): 53–62. PubMed Abstract \| Publisher Full Text\n\nStoll BJ, Hansen NI, Bell EF, et al.: Neonatal outcomes of extremely preterm infants from the NICHD Neonatal Research Network. Pediatrics. 2010; 126(3): 443–56. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStolp HB, Dziegielewska KM, Ek CJ, et al.: Long-term changes in blood-brain barrier permeability and white matter following prolonged systemic inflammation in early development in the rat. Eur J Neurosci. 2005; 22(11): 2805–2816. PubMed Abstract \| Publisher Full Text\n\nStrazielle N, Ghersi-Egea JF: Efflux transporters in blood-brain interfaces of the developing brain. Front Neurosci. 2015; 9: 21. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStuddert V, Gay C, Blood D: Saunders Comprehensive Veterinary Dictionary. 4th Edition, Elsevier, Amsterdam, NL. 2011. Reference Source\n\nTaskar KS, Mariappan TT, Kurawattimath V, et al.: Unmasking the Role of Uptake Transporters for Digoxin Uptake Across the Barriers of the Central Nervous System in Rat. J Cent Nerv Syst Dis. 2017; 9: 1179573517693596. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWerler MM, Mitchell AA, Hernandez-Diaz S, et al.: Use of over-the-counter medications during pregnancy. Am J Obstet Gynecol. 2005; 193(3 Pt 1): 771–777. PubMed Abstract \| Publisher Full Text\n\nWorld Health Organisation: WHO guidelines on the pharmacological treatment of persisting pain in children with medical illnesses. WHO Press, Geneva. 2012. PubMed Abstract\n\nWyszynski DF, Shields KE: Frequency and type of medications and vaccines used during pregnancy. Obstet Med. 2016; 9(1): 21–27. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "52190", "date": "12 Aug 2019", "name": "Alan Leviton", "expertise": [ "Reviewer Expertise epidemiology of brain disorders in the extremely preterm newborn." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAs I scanned this manuscript (ms), I thought the introduction rather long. On reading it, however, I was impressed how well it laid out the field succinctly, identified some of what was missing, how this report would eliminate one deficiency, and then what was found. Well done, indeed. The design is more than appropriate with the technically-sound assessment of multiple drugs and multiple passive permeability markers\n\nThe methods and analysis are presented in sufficient detail for others to attempt replication. I was at first frustrated that I could not find the number of animals identified in the text. Then I found Tables 1 and 2, which, on consideration, seemed eminently reasonable given the numbers that were appropriately listed for each age group for each assessment. The authors used Tables 3 and 4 to identify materials used. Yes, the authors provided all the information needed.\n\nThe 4 data tables (#s 5, 6, 7, and 8) and the 7 figures present all the data needed very well.\n\nThe authors clearly document changes to support their conclusion that the functional capacities of efflux transporters at the barriers are developmentally regulated.\n\nI did NOT think that any revisions are called for.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "52902", "date": "09 Sep 2019", "name": "Reina Bendayan", "expertise": [ "Reviewer Expertise Drug Transport across Biological Membranes and Blood-Tissue Barriers" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nSummary-Study Objectives:\n\nThis is a very interesting and unique study performed by an excellent group of researchers with high expertise in the field. Overall, the study provides novel data describing the permeability of three commonly used drugs in the clinic, digoxin, cimetidine and paracetamol, into the blood, brain and CSF compartments of Sprague Dawley rats at three developmental stages. These data are further complemented by qPCR-based RNA expression studies for relevant ABC transporters in the placenta, brain and choroid plexus. The findings highlight a link between novel age-specific differences in transporter regulation and penetration of these drugs into blood, brain and CSF compartments. The work significantly advances the knowledge in the field as there is a very important lack of information related to this topic.\nOverall Comments:\nThe supporting background information described in the introduction and discussion is a very thorough representation of existing literature, and gives context to the results. Overall, the studies are carefully planned and the methods applied are robust. Limitations of the study are thoroughly described in the discussion. This work sets the stage for further investigation into the topic, and future directions are highlighted by the authors. The work significantly advances the knowledge in the field as there is a very important lack of information related to this topic.\nSpecific Comments:\n\nAdditional details on the drug dosing would be helpful especially since there is ample literature describing the use of the drugs in animal models. It is clear that the dosing was adapted from the Australian Medicines Handbook; however, the authors should clarify if allometric scaling or existing literature was used to adapt the drug doses for rodent studies. The authors should also address whether the observed drug plasma concentrations in their animal model reflect clinical relevant plasma concentrations.\n\nThe authors should consider the extent of drug plasma protein binding when they address the various factors regulating drug permeability across the various tissue compartments; only free drug will be able to permeate across biological membranes by diffusion.\n\nDrug concentration in biological compartments is the result of the interplay of passive diffusion across biological membranes, influx/efflux transport, and drug metabolism. While the authors have nicely characterized the role of efflux transporters, they have not fully addressed the potential role of influx transporters in the permeability of the drugs studied. Members of the Solute Carrier (SLC) superfamily could be involved in the uptake of the drugs across the various compartments. The expression of these transporters could also be very variable at the different development stages. This could be further addressed in the discussion for completeness.\n\nMeasurement of radiolabeled drug in the tissue compartments will reflect total drug/metabolite concentrations rather than parent drug; hence drug metabolites may present a confounding factor in the interpretation of the data. This should be addressed by the authors.\n\nIn the discussion section, p17 second paragraph, please correct the spelling of the terms “glucuronidation” “glutathionylation”.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "52901", "date": "18 Sep 2019", "name": "Richard Day", "expertise": [ "Reviewer Expertise Clinical Pharmacologist", "Rheumatologist" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe rationale for this study is the potential for transfer of medicines into the fetal and infant brain, a subject that is little understood through lack of research. It is a relevant question as there is considerable exposure of the undeveloped brain to medicines via administration to pregnant women, breast feeding women and directly to infants. The focus is upon the ABC transporter family, efficient at protecting the adult brain from lipid soluble medicines, but about which little is known in the developing brain. This question is addressed by using labelled digoxin, cimetidine and paracetamol, substrates for 3 different ABC transporters. Transport into brain and CSF is measured using radioactively labelled drug, dosed acutely or chronically and at different stages of brain development. Expression patterns of the transporters are examined at different stages of animal development and correlations sought with the transport of drugs into brain/CSF.\n\nThere were large contrasts between E19 animals and P4/Adults for brain/plasma and CSF/plasma ratios for all three drugs, (but perhaps less so for paracetamol, especially acute administration) with E19 having greater transport/permeability. Useful information regarding ratios of marker compounds with a range of octanol-water coefficients, drug ratios and transporter expression for acute and chronic treatment was obtained. The ‘system’ showed potential for investigation of new medicines, prioritising studies of drug action on brain development on the basis of uptake into brain, e.g. paracetamol’s greater propensity for transport into brain at E19, unravelling protective effects of transporters and adaptions to componentry of the ‘system’ with dosage, duration of therapy and drug combinations. The work is well done, clearly described, well illustrated and fairly presented.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1372
https://f1000research.com/articles/7-1918/v1	10 Dec 18	{ "type": "Research Article", "title": "Chicken feather hydrolysate as alternative peptone source for microbial cultivation", "authors": [ "Oghenerobor B. Akpor", "Damilola E. Odesola", "Remilekun E. Thomas", "Olarewaju M. Oluba", "Damilola E. Odesola", "Remilekun E. Thomas", "Olarewaju M. Oluba" ], "abstract": "Background: Commercially available conventional growth medium for the culture of microbes are expensive, hence the need for alternative cheaper sources. Poultry waste, in the form of feather and blood, are of value in biotechnology because of their high protein content. Hence the primary aim of this study was to produce a cheaper peptone alternative from chicken feather protein hydrolysate (CFPH) and blood meal (BM). Methods: We monitored the growth of selected bacteria and fungi in different concentrations of medium produced from varying combination of peptone, CFPH and BM in order to determine the combination that produced maximum growth. Five different media, namely 100% peptone (control), 100% BM, 40% peptone + 60% CFPH, 40% BM + 60% CFPH and 20% peptone + 20% BM + 60% CFPH were prepared and used for the study. The different media were inoculated with 1 ml of each test organism (Escherichia coli, Klebsiella pneumoniae, Proteus mirabilis, Staphylococcus aureus, Pseudomonas aeruginosa, Candida carpophila, Candida tropicalis and Pichia kundriavzevii) and their growth monitored for 10 h. Results: Pseudomonas aeruginosa, Proteus mirabilis and Staphylococcus aureus grew best in the 100% peptone, Klebsiella pneumoniae grew best in 100 BM. The fungi species were observed to grow best in 100% peptone. The 60% CFPH + 40% peptone combination (CFPH obtained with precipitate of TCA, HCl, and HNO3 gave the best growth of E. coli. The 60% CFPH + 40% peptone combination (CFPH obtained with precipitate of TCA) also gave the best growth of C. tropicalis and Klebsiella pneumoniae respectively. Conclusions: Overall, the 60% CFPH + 40% peptone combination showed the most potential as an alternative to peptone, especially for E. coli.", "keywords": [ "Culture media", "chicken feather", "keratin", "hydrolysate", "bacteria", "yeasts", "growth rate", "protein source" ], "content": "Introduction\n\nMicrobial culture medium is composed of different nutrients required by organisms for growth. The nutrient requirements of microorganisms differ from one to another as there are many types of microorganisms. Generally, the microbial growth composition includes carbon and energy sources, protein hydrolysates, otherwise known as peptones, extracts, buffers and sometimes gelling agents. Microorganisms will only be able to grow if they are provided with the appropriate nutrients for growth. Apart from carbon, microbes require a source of nitrogen. Amino acids, urea, ammonia and other compound may serve as the nitrogen source. Some organisms also possess the ability to metabolize peptides and more complex proteins (Sandle, 2016).\n\nThe protein source for microbial culture is derived from peptone, which is a good source of amino acids, peptides and proteins in growth medium. It is an excellent source of nitrogen. However, it is also a very expensive constituent of the microbial culture medium (Taskin & Kubanoglu, 2011). Different natural products, such as milk, animal tissues and plants, are being exploited for obtaining peptone in order to reduce the costs of production of the growth medium (Jayathilakan et al., 2012; Ben Rebah & Miled, 2013).\n\nIt is estimated that about 20 million tons of chicken feathers are generated weekly worldwide and are considered menaces in terms of solid waste pollution (Egelyng et al., 2018; Tesfaye et al., 2018). In a bid to dispose of the large amounts waste chicken feathers generated worldwide, methods such as landfill and burning have been used, which have taken a toll on the environment (Sharma et al., 2017).\n\nTo reduce the burden of waste generated by disposed chicken feathers, a number of processes and operations involving the application of chicken feathers have been reported. Chicken feathers have been employed in the production of animal feed, textile production and paper production amongst many others (Chinta et al., 2013; Moritz & Latshaw, 2001; Tesfaye et al., 2017).\n\nChicken feathers contain more than 90% protein (keratin), 1% lipids and 8% water (Lasekan et al., 2013). Keratin proteins are grouped into the alpha and beta keratins. Chicken feathers and feathers from most birds are composed majorly of the beta-keratin. Keratin contains all 20 amino acids linked together by peptide bonds, which include covalent disulphide bonds, ionic bonds, hydrogen bonds and hydrophobic bonds (Greenwold et al., 2014). Keratins are bonded by a number of these bonds which make them naturally insoluble. These bonds require that they be broken in other to obtain the chicken feathers in usable forms for microorganisms. By hydrolysis, the bonds are broken, a soluble product is formed (hydrolysate) and the peptone can be obtained from the chicken feather keratin (Ayutthaya & Wootthikanokkhan, 2013). The conversion of such large amounts of chicken feathers into hydrolyzed forms for the manufacture of microbial culture medium can be used as a measure of solving this problem. The chicken feather keratin can then be incorporated into the production of microbial culture medium. This study was therefore targeted at utilizing the keratin in the waste chicken feathers as a cheaper alternative to peptone and also a nitrogen source for microbial growth.\n\n\nMethods\n\nChicken feathers were collected from the poultry house in Landmark University Commercial Farm in Omu-Aran, Kwara State, Nigeria. The feathers were first washed with water and laundry detergent before disinfecting with 5% hypochloride solution. Following disinfection, the feathers were sun-dried for one week and stored in baskets until when needed.\n\nFor hydrolysis, approximately 400 g of dried chicken feathers were placed in 10-l plastic container, after which 2000 ml 1 M NaOH was added. The feather-NaOH mixture was stirred vigorously and left to stand for 10 h. After hydrolysis, the mixture was filtered with a clean, dry muslin cloth and the unhydrolyzed fraction estimated. The quantification of the unhydrolyzed fraction was carried out after drying in an oven to constant weight and then weighed to ascertain the degree of hydrolysis.\n\nFeather keratin was precipitated separately from the hydrolyzed feather solution, using 1 M solutions of the following acids: HCl, H2SO4, C2HCl3O2 (TCA) and HNO3. This was carried out for 10 min at 25°C.\n\nThe precipitated chicken feather hydrolysate was separated from the solution by filtration. The chicken feather hydrolysate of the respective acids was air-dried and quantified. The chicken feather hydrolysate of the respective acids (henceforth referred to as feather keratin or hydrolysate) were ground into fine powders using a laboratory blender.\n\nFresh cattle blood was obtained from the abattoir of Landmark University, Omu-Aran (Nigeria) Teaching and Research Farm. The blood was immediately heated at 60°CC for 30 min to coagulate. The coagulated blood was then placed in aluminium foil and oven-dried at 50°C for 8 h. The dried coagulated blood was milled into powder, using a mechanical grinder. The powdered blood, referred to as blood meal (BM) was kept in air-tight container until required.\n\nFor this study, different compositions of feather keratin, peptone (Oxoid, UK) and BM were used in the medium formulations. Five different combinations were used to make the respective growth medium, as follows. Combination A: 60% keratin, 40% peptone, 0% BM; Combination B: 60% keratin, 20% peptone, 20% BM; Combination C: 60% keratin, 0% peptone, 40% BM; Combination D: 100% peptone; Combination E: 100% BM.\n\nThe respective components were weighed separately and dissolved in aliquot quantities of distilled water before combining all the components together to make a final medium concentration of 5 g/l. Each medium, as composed, was dispensed in conical flask and sterilized in an autoclave (121°C for 15 min) at 15 psi.\n\nA total of eight microbial species, consisting of five bacteria and three yeast species were used for the study. The bacteria species were Escherichia coli, Klebsiella pneumoniae, Proteus mirabilis, Staphylococcus aureus and Pseudomonas aeruginosa. The yeasts were Candida carpophila, Candida tropicalis and Pichia kundriavzevii.\n\nPrior to use, stock microbial cultures of the isolates were first streaked on agar plates to ascertain their purity before subculturing into nutrient broth. Equal volumes of broth cultures of the pure isolates (0.5 ml) were used for inoculation. Bacteria and fungi were incubated at 24±2°C.\n\nFor growth studies, the flasks containing the sterile medium were inoculated with the test microbial species. Following inoculation and every 2 h, for a 10 h duration, aliquot samples were withdrawn from each flask to monitor growth by measuring absorbance using a spectrophotometer at a wavelength of 750 nm, using sterile nutrient broth to normalize. Each experiment was carried out in duplicate.\n\nData analysis was carried out using the SPSS statistical software, version 13.0. Comparison of means was determined using the one-way ANOVA test at a significance level of P<0.05. For the least-significant-difference post hoc multiple comparison test was used. All experimental setups were in duplicates.\n\n\nResults\n\nRaw absorbance values for each microbe are available on figshare (Akpor et al., 2018).\n\nAs shown in Figure 1, the growth pattern of Candida carpophila in all the acid hydrolysate medium showed consistent increase with time. The highest growth was observed in the medium with 100% peptone. This trend was irrespective of the acid hydrolysate used. Besides, the 100% peptone medium, the ranking of the growth of the organism from the highest to the least in the other medium compositions varied for each acid feather hydrolysate. In the TCA feather hydrolysate medium, after 10 hours, the organism recorded the highest growth in the 100% peptone medium, closely followed by 60% feather hydrolysate + 20% peptone + 20% blood, 60% feather hydrolysate + 40% peptone, and 100% blood in that order, while the least growth was observed with 60% feather hydrolysate + 40% blood (Figure 1).\n\nFor the HCl feather hydrolysate medium, the growth pattern observed followed the order: 100% peptone > 100% blood > 60% feather hydrolysate + 20% peptone + 20% blood > 60% feather hydrolysate + 40% peptone, while the 60% feather hydrolysate + 40% blood medium showed the least growth. When H2SO4 feather hydrolysate medium was used, the growth was in the order: 100% peptone > 100% blood > 60% feather hydrolysate + 40% blood > 60% feather hydrolysate + 20% peptone + 20% blood, while the 60% feather hydrolysate + 40% peptone was least. In the HNO3 feather hydrolysate medium, C. carpophila grew in the order 100% peptone > 100% blood > 60% feather hydrolysate + 20% peptone + 20% blood > 60% feather hydrolysate + 40% blood, and the lowest growth was recorded in the 60% feather hydrolysate + 40% peptone (Figure 1).\n\nAs represented in Figure 2, the highest and the lowest growth of Candida tropicalis varied with medium composition for each of the acid feather hydrolysates used. However, the most frequent medium composition in which the highest growth of the organism was recorded was 100% peptone and the organism showed consistent increase in growth throughout the 10-h period with the highest growth at the 10th hour. When the trichloroacetic acid feather hydrolysate medium was used, C. tropicalis showed no significant difference (p > 0.05) in growth with both 60% feather hydrolysate+ 20% peptone + 20% blood and 60% feather hydrolysate + 40% peptone and the organism’s growth was observed to be highest with both compositions. This was followed by 100% peptone, 60% feather hydrolysate + 40% blood and the least growth was with 100% blood (Figure 2).\n\nIn the HCl feather hydrolysate medium, the growth of C. tropicalis was highest with 100% peptone followed by 60% feather hydrolysate + 40% blood, 60% feather hydrolysate + 20% peptone + 20% blood, 60 feather hydrolysate+ 40% peptone, with the lowest growth occurring in the 100% blood medium. For the H2SO4 feather hydrolysate medium, the highest growth of C. tropicalis was observed with 100% peptone, followed by 100% blood, 60% feather hydrolysate + 40% blood, 60% feather hydrolysate + 20% peptone + 20% blood and the least growth was observed with 60% feather hydrolysate+ 40% peptone. In the HNO3 hydrolysate medium, C. tropicalis had its highest growth with 100% peptone, followed by 100% blood, 60% feather hydrolysate + 40% peptone, 60% feather hydrolysate + 40% blood, and had its least growth with 60% feather hydrolysate + 20% peptone + 20% blood (Figure 2).\n\nAs illustrated in Figure 3, the E. coli showed consistent increase in growth with the highest growth at the 10th hour. The highest and lowest growths of the organism varied with medium composition for all the acid feather hydrolysates. In the TCA feather hydrolysate medium, the highest growth of E. coli at the end of the 10 h period was observed with 60% feather hydrolysate + 40% peptone, followed by 60% feather hydrolysate + 20% peptone + 20% blood, 100% peptone while the least growth was observed with 100% blood and 60% feather hydrolysate + 40% blood. There was no significant difference (p > 0.05) in growth of the organism in both medium compositions. When the HCl feather hydrolysate medium was used, highest growth of the E. coli was with 100% peptone, followed by 60% feather hydrolysate+ 40% peptone, 60% feather hydrolysate + 20% peptone + 20% blood and 60% feather hydrolysate + 40% blood while the least growth was with 100% blood (Figure 3).\n\nFor the H2SO4 feather hydrolysate medium, E. coli had its highest growth with 100% peptone followed by 60% feather hydrolysate + 40% peptone, and then 60% feather hydrolysate+ 40% blood and 60% feather hydrolysate + 20% peptone + 20% blood in which there was little or no significant difference in growth while the least growth of E. coli was recorded with 100% peptone. In the HNO3 feather hydrolysate medium, the highest growth of E. coli recorded was with 60% feather hydrolysate+ 40% peptone, followed by 100% peptone, while the least growths were observed with 60% feather hydrolysate + 40% blood, 60% feather hydrolysate + 20% peptone + 20% blood and 100% blood, all showing no significant difference (p > 0.05) in growth (Figure 3).\n\nThe results in Figure 4 showed consistent increase in growth rate of Klebsiella pneumoniae throughout the 10 h period with highest growth at the 10th hour. The highest and lowest growths of the organism varied with the medium composition. In the trichloroacetic acid feather hydrolysate medium, the highest growth of Klebsiella pneumoniae was with 100% blood, followed by 60% feather hydrolysate + 40% peptone, 60% feather hydrolysate + 40% blood, 100% peptone, and the least growth was with 60 feather hydrolysate + 20% peptone + 20% blood. In the HCl feather hydrolysate medium, the Klebsiella pneumoniae had its highest growth with 100% blood, followed by 60% feather hydrolysate+ 40% peptone, 60% feather hydrolysate + 40% blood, 60% feather hydrolysate + 20% peptone + 20% blood, and the least growth was with 100% peptone (Figure 4).\n\nIn the H2SO4 feather hydrolysate medium, the medium composition in which Klebsiella pneumoniae had its highest growth was 100% blood, followed by 60% feather hydrolysate + 40% blood, 60% feather hydrolysate + 40% peptone and then, 60% feather hydrolysate + 20% peptone + 20% blood, while the least was recorded in 100% peptone medium. In the HNO3 feather hydrolysate medium, the highest growth of Klebsiella was observed in 100% blood, followed by 60% feather hydrolysate + 40% peptone, 60% feather hydrolysate + 40% blood, 60% feather hydrolysate+ 20% peptone + 20% blood and the least growth of the organism was observed in 100% peptone (Figure 4).\n\nAs represented in Figure 5, Pseudomonas aeruginosa showed consistent increase in growth for the 10 h period and the highest growth was recorded at the 10th hour. The medium composition which yielded the highest growth of the organism was 100% peptone and the least growth was recorded with 100% blood medium. This was constant for all the acid hydrolysate medium. In the TCA feather hydrolysate medium, the medium composition after 100% peptone that yielded the highest growth of P. aeruginosa was 60% feather hydrolysate+ 40% peptone, followed by 60% feather hydrolysate + 20% peptone + 20% blood, 60% feather hydrolysate + 40% blood, while the least growth of Pseudomonas aeruginosa was with 100% blood (Figure 5).\n\nIn the HCl feather hydrolysate medium, after 100% peptone, P. aeruginosa had its highest growth with 60% feather hydrolysate + 40% blood, followed by 60% feather hydrolysate + 40% peptone, 60% feather hydrolysate + 20% peptone + 20% blood, while the organism had its least growth with 100% blood. In the H2SO4 feather hydrolysate medium, 100% peptone remained the medium composition in which the highest growth of P. aeruginosa was recorded, followed by 60% feather hydrolysate + 40% blood, 60% feather hydrolysate + 40% peptone and 60% feather hydrolysate + 20% peptone + 20% blood while the least growth was observed in 100% blood. In the HNO3 feather hydrolysate medium, following 100% peptone, the highest growth of P. aeruginosa was observed with 60% feather hydrolysate+ 40% blood, followed by 60% feather hydrolysate + 40% peptone, 60% feather hydrolysate + 20% peptone + 20% blood. The least growth was observed with 100% blood medium (Figure 5).\n\nFigure 6 shows the growth pattern of Pichia kudriavzevii in the different medium. Pichia kudriavzevii showed consistent increase in growth with the highest growth at the 10th hour. The highest growth of the organism was 100% peptone medium, while the least growth was observed in 100% blood medium. In the TCA feather hydrolysate medium, the order of growth of Pichia kudriavzevii from the highest to the least in the medium compositions was 100% peptone > 60% feather hydrolysate + 40% peptone > 60% feather hydrolysate + 40% blood > 60% feather hydrolysate, + 20% peptone + 20% blood >100% blood (Figure 6).\n\nIn the HCl feather hydrolysate medium, the order of growth was: 100% peptone > 60% keratin + 40% peptone > 60% feather hydrolysate + 40% blood > 60% feather hydrolysate + 20% peptone + 20% blood > 100% blood. In the H2SO4 feather hydrolysate medium, the growth of the organism followed the order: 100% peptone > 60% feather hydrolysate + 40% blood > 60% feather hydrolysate + 40% peptone > 60% feather hydrolysate + 20% peptone + 20% > 100% blood. In the HNO3 feather hydrolysate medium, the order of growth in the medium compositions was: 100% peptone > 60% feather hydrolysate + 40% blood > 60% feather hydrolysate + 40% peptone > 60% feather hydrolysate + 20% peptone + 20% blood > 100% blood (Figure 6).\n\nThe growth rate of the Proteus mirabilis in the different medium is shown in Figure 7. As shown in the Figure, the highest growth was recorded at the 10th hour for all the medium compositions against the respective acid hydrolysates\n\nIn the TCA feather hydrolysate medium, besides 100% peptone, in which the highest growth was recorded, the second highest growth of Proteus mirabilis was with 100% blood medium, followed by 60% feather hydrolysate + 40% peptone, 60% feather hydrolysate + 40% blood and the least growth was with 60% feather hydrolysate + 20% peptone + 20% blood. In the HCl feather hydrolysate medium, after 100% peptone, the highest growth of Proteus mirabilis was observed with 100% blood, followed by 60% feather hydrolysate + 40% blood, 60% feather hydrolysate + 20% peptone + 20% blood, and the least growth was with 60% feather hydrolysate + 40% peptone (Figure 7).\n\nIn the H2SO4 feather hydrolysate medium, after 100% peptone, Proteus mirabilis growth pattern followed the order 100% peptone > 60% feather hydrolysate + 40% peptone > 60% feather hydrolysate + 40% blood 60% feather hydrolysate + 20% peptone + 20% blood (Figure 7).\n\nIn presence of the different hydrolysates, Staphylococcus aureus showed consistent increase in growth throughout the period of incubation. The 100% peptone medium produced the best growth against all the acid hydrolysate medium, while the least growth of the organism varied with medium composition.\n\nIn the TCA feather hydrolysate medium, the 60% feather hydrolysate + 40% peptone medium was second to the 100% peptone medium in supporting the growth of Staphylococcus aureus, followed by 60% feather hydrolysate + 20% peptone + 20% blood, 100% blood and 60% feather hydrolysate + 40% blood medium in that order. In the HCl feather hydrolysate medium, the second highest growth of the Staphylococcus aureus was observed with 60% feather hydrolysate + 40% blood, followed by 100% blood, 60% feather hydrolysate + 40% peptone, and 60% feather hydrolysate + 20% peptone + 20% blood in that order (Figure 8).\n\nIn the H2SO4 feather hydrolysate medium, the 60% feather hydrolysate+ 20% peptone + 20% blood medium ranked second to the 100% peptone medium in the observed growth pattern for Staphylococcus aureus, followed by 60% feather hydrolysate + 40% blood, 60% feather hydrolysate + 40% peptone, and 100% blood medium in that order. In the nitric acid feather hydrolysate medium, the growth pattern observed for Staphylococcus aureus followed the order 100% peptone > 60% feather hydrolysate + 40% blood > 60% feather hydrolysate + 40% peptone > 60% feather hydrolysate + 20% peptone + 20% blood (Figure 8).\n\n\nDiscussion\n\nIn this study, the chicken feather keratin peptone was obtained by alkaline hydrolysis and acid neutralization and precipitation delete. In most cases KOH, NaOH and Ca(OH)2 are used for the hydrolysis. Taskin et al., (2016) used the alkaline hydrolysis method with KOH in a study where peptone was obtained from sheep wool protein hydrolysate. From investigations, alkaline hydrolysis was reported to be able to produce a high yield of keratin and also enhance the keratin extraction effectiveness (Sinkiewicz et al., 2017). Alkaline hydrolysis has also been studied and proven to be effective in the degradation of waste containing keratin and collagen (Gousterova et al., 2005). It is opined that the use of alkaline hydrolysis inactivates of pathogens and prions such as transmissible spongiform encephalopathy (TSE) and bovine spongiform encephalopathy (BSE). The use of alkaline hydrolysis yields a BSE and TSE free hydrolysate medium (Matthews & Cooke, 2003). Despite the fact that heat is employed in most chemical hydrolysis processes to improve yield, the alkaline hydrolysis in this study was carried out under room temperature for approximately 10 h. Chemical hydrolysis conducted at high temperatures is said to lead to the destruction of amino acids (Sinkiewicz et al., 2017).\n\nAlthough acid hydrolysis has also been used in some studies to obtain protein hydrolysates (Wisuthiphaet & Kongruang, 2015), it is argued that it could result in destruction of essential amino acids, such as methionine, cystine, cysteine and tryptophan, and conversion of glutamine and aspargine to glutamic and aspartic acid, respectively. It is indicated that during acid hydrolysis, the salts in hydrolysates can be injurious to the growth of the microorganisms (Bucci & Unlu, 2000).\n\nThis study aimed to assess a cheaper source of protein for microbial culture than conventional nutrient medium. Chicken feathers were chosen as the material for research because of its abundance, cost-efficiency and high protein content. The results obtained in this study with reference to the performance of the organisms in the formulated medium compositions were viewed in comparison with the performance of the organisms in the commercially produced peptone. The comparison of the growth rate of the organisms in the different medium with that in peptone was deliberate. Studies have revealed peptone as an excellent nitrogen and protein source for growing microorganisms and manufacture of growth medium (Markings, 2018).\n\nThe results of the present experiment showed that in most cases, the organisms growth was highest with the 100% commercially produced peptone while in other cases, various medium compositions yielded higher growth of the organisms even better than in the 100% peptone.\n\nThe medium compositions that favored the highest growth of the organism differed with each acid hydrolysate for every organism. The best growth of Candida carpophila was recorded with the commercial peptone irrespective of the acid hydrolysate medium composition used. However, in the case of Candida tropicalis, the formulated medium compositions from the chicken feather hydrolysate were able to yield higher growths of Candida tropicalis even better than the commercial peptone but only in the TCA hydrolysate, while the commercial peptone yielded the highest growth of the organisms in the other acid hydrolysates. Pseudomonas aeruginosa, Pichia Kudriavzevii, Proteus mirabilis and Staphylococcus aureus consistently had their highest growth with the commercial peptone irrespective of the acid hydrolysate used. In Escherichia coli and Klebsiella pneumoniae, variation of the highest growth with the hydrolysate medium compositions and growth yield of the organisms higher than in the commercial peptones persisted. The variations in the medium compositions that showed the best growth of the organisms were indications that the acids used in precipitating the keratin from the chicken feathers had significant impact on the efficacy of the medium compositions. The optimum concentration of chicken feather keratin concentration that yielded maximum microbial growth was 60% feather hydrolysate + 40% peptone in Escherichia coli.\n\nThe ideal chicken feather hydrolysate that yielded microbial growth was the TCA hydrolysate. In a study by Taskin et al. (2016), using sheep wool protein hydrolysate as a peptone source for microorganisms, Staphylococcus aureus was observed to show poor growth performance in the medium. However, growth performances of Saccharomyces cerevisiae, Bacillus subtilis and Penicillium chrysogenum were observed to be moderate in wool protein. Generally, the present study revealed that growth rate of the respective organisms varied from one medium composition to the other. This observation has been reported by earlier workers (Aspmo et al., 2005; Jones & Fayerman, 1987).\n\n\nConclusion\n\nWaste chicken feathers that were previously discarded and viewed as a burden to the environment can now be viewed as an important bio-resource and can be manipulated and explored widely as a biotechnological material. This study was able to characterize waste chicken feathers as a microbiological tool for microbial culture. Because of its high protein content, protein essential for the growth of microorganisms can be synthesized from chicken feathers.\n\nIn the study, the potential of the chicken feather keratin to support the growth of both bacteria and fungi was established. Chicken feather hydrolysate was also proven to be an excellent substrate particularly for the growth of Escherichia coli. Using a suitable process of hydrolysis, peptone from chicken feather keratin can be re-modified, industrialized and produced in bulk on a commercial scale. This research work could serve as precursor to exploring many other waste materials of high protein content which could be of biotechnological value.\n\n\nData availability\n\nThe raw absorbance values from analysis of microbe growth using different growth media are available on figshare. DOI: https://doi.org/10.6084/m9.figshare.7376564.v1 (Akpor et al., 2018).\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nAkpor O, Oluba O, Odesola D, et al.: Chicken feather hydrolysate as alternative peptone source for microbial cultivation. figshare. Dataset. 2018.\n\nAspmo SI, Horn SJ, Eijsink GHV: Hydrolysates from Atlantic Cod (Gadus morhua L.) viscera as components of microbial growth media. Process Biochem. 2005; 40(12): 3714–3722. Publisher Full Text\n\nAyutthaya SIN, Wootthikanokkhan J: Extraction of Keratin from Chicken Feather and Electrospinning of the Keratin/PLA Blends. Adv Mat Res. 2013; 747: 711–714. Publisher Full Text\n\nBen Rebah F, Miled N: Fish processing wastes for microbial enzyme production: a Review. 3 Biotech. 2013; 3(4): 255–265. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBucci LR, Unlu L: Protein and amino acid supplements in exercise and sport. In Wolinsky I, Driskell JA, Energy-yielding macronutrients and energy metabolism in sports nutrition. Boca Raton, Florida: CRC Press. 2000; 191–212. Reference Source\n\nChinta SK, Landage SM, Yadav K: Application of chicken feathers in technical textiles. International Journal of Innovative Research in Science, Engineering and Technology. 2013; 2(4): 1158–1165. Reference Source\n\nEgelyng H, Romsdal A, Hansen HO, et al.: Cascading Norwegian co-streams for bioeconomic transition. J Clean Prod. 2018; 172: 3864–3873. Publisher Full Text\n\nGousterova A, Braikova D, Goshev I, et al.: Degradation of keratin and collagen containing wastes by newly isolated thermoactinomycetes or by alkaline hydrolysis. Lett Appl Microbiol. 2005; 40(5): 335–340. PubMed Abstract \| Publisher Full Text\n\nGreenwold MJ, Bao W, Jarvis ED, et al.: Dynamic evolution of the alpha (α) and beta (β) keratins has accompanied integument diversification and the adaptation of birds into novel lifestyles. BMC Evol Biol. 2014; 14: 249–264. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJayathilakan K, Sultana K, Radhakrishna K, et al.: Utilization of byproducts and waste materials from meat, poultry and fish processing industries: a review. J Food Sci Technol. 2012; 49(3): 278–293. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJones ME, Fayerman JT: Industrial applications of recombinant DNA technology. J Chem Educ. 1987; 64: 337–339. Publisher Full Text\n\nLasekan A, Bakar AF, Hashim D: Potential of chicken by-products as sources of useful biological resources. Waste Manag. 2013; 33(3): 552–565. PubMed Abstract \| Publisher Full Text\n\nMarkings S: The chemical composition of nutrient agar. 2018; Reference Source\n\nMatthews D, Cooke BC: The potential for transmissible spongiform encephalopathies in non-ruminant livestock and fish. Rev Sci Tech. 2003; 22(1): 283–296. PubMed Abstract \| Publisher Full Text\n\nMoritz JS, Latshaw JD: Indicators of nutritional value of hydrolyzed feather meal. Poult Sci. 2001; 80(1): 79–86. PubMed Abstract \| Publisher Full Text\n\nSandle T: Pharmaceutical Microbiology Essentials for Quality Assurance and Quality Control. Cambridge: Woodhead Publishing. 2016. Reference Source\n\nSharma S, Gupta A, Chik SM, et al.: Dissolution and characterization of biofunctional keratin particles extracted from chicken feathers. IOP Conf Ser Mater Sci Eng. 2017; 191: 1–5. Publisher Full Text\n\nSinkiewicz I, Siliwinska A, Staroszczynk H, et al.: Alternative methods of preparation of soluble keratin from chicken feathers. Waste Biomass Valorization. 2017; 8(4): 1043–1048. Publisher Full Text\n\nTaskin M, Kubanoglu EB: Evaluation of waste chicken feathers as peptone source for bacterial growth. J Appl Microbiol. 2011; 111(4): 826–834. PubMed Abstract \| Publisher Full Text\n\nTaskin M, Unver Y, Firat A, et al.: Sheep wool protein hydrolysate: a new peptone source for microorganisms. J Chem Technol Biotechnol. 2016; 91(6): 1675–1680. Publisher Full Text\n\nTesfaye T, Sithole B, Ramjugernath D, et al.: Valorisation of chicken feathers: Application in paper production. J Clean Prod. 2017; 164: 1324–1331. Publisher Full Text\n\nTesfaye T, Sithole B, Ramjugernath D: Valorisation of chicken feather barbs: Utilisation in yarn production and technical textile applications. Sustainable Chem Pharm. 2018; 8: 38–49. Publisher Full Text\n\nWisuthiphaet N, Kongruang S: Production of fish protein hydrolysates by acid and enzymatic hydrolysis. Journal of Medical and Bioengineering. 2015; 4(6): 466–470. Publisher Full Text" }	[ { "id": "42168", "date": "22 Jan 2019", "name": "Swati Sharma", "expertise": [ "Reviewer Expertise biomaterials", "biopolymers", "proteins", "waste management" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAbstract\nGive full forms of TCA... .....also gave the best growth of C. tropicalis and Klebsiella pneumoniae respectively. In this sentence in what respect is respectively used? In the manuscript total eight microorganisms were considered and conclusion is all about one microorganism. Justify? Specify from where blood meal was taken.\nIntroduction\nKeratin proteins are grouped into the alpha and beta keratins. Chicken feathers and feathers from most birds are composed majorly of the beta keratin. Provide with suitable reference.\nMethods\nCleaning method of feathers was obtained from previous research. Need to cite the reference for the same What was the basis to choose 1M NaOH for the study. Why didn't you try different concentrations? Is it standardized in your laboratory or taken from previous research, if so cite the reference. On what basis were acids chosen for precipitation? Heated at 60°CC for...... remove double C. Naming should be same for the combination throughout the manuscript Why didn't you try 100% feather hydrolysate? Correct the mistake in sentence ed by 60% feather hydrolysate + 20% peptone + 20% blood, 60% feather hydrolysate + 40% peptone, and 100% blood in that order, while the least growth was observed with 60% feather hydrolysate + 40% blood....it should be Blood meal not blood in whole manuscript and same thing apply for feather hydrolysate Give names to all medium like a, b...\nResults\nOverall the result section is good. Conclusion is well written.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4373", "date": "30 Jan 2019", "name": "Oghenerobor Akpor", "role": "Author Response", "response": "The authors appreciate thoroughness of the reviewer and the comments. Almost all the points raised that required modification have been attended to in the revised version. With respect to the use 1 M NaOH and the acids for hydrolysis and precipitation, respectively, that was ascertained during method development and the pilot phase of the study." } ] }, { "id": "42593", "date": "29 Jan 2019", "name": "Adriano Brandelli", "expertise": [ "Reviewer Expertise Microbial biotechnology" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nPoultry industry experienced a constant growth during the last decades, generating an increased amount of waste, in the form or feathers, viscera, skin, and dead-on-arrival (Brandelli et al. 20151). Alternatives for the correct management of such poultry waste have been evaluated. Conversion of feathers into hydrolysates is a possible alternative to produce ingredients for feed, fertilizers, and cultivation media formulations (Daroit and Brandelli 20142). This topic merits investigation, since the correct management of such waste material has great environmental relevance. In addition, it could represent an added value to the industry.\nIn the current article, feathers were hydrolyzed by chemical methods and the resulting hydrolysates were tested as a peptone source for microbial growth. The utilization of byproducts obtained from abattoir as ingredients for culture media has been previously described. In particular, waste feathers from poultry processing industry have been converted as peptone source for bacterial growth (Taskin and Kurbanoglu, 20113) and for development of culture media for production of mosquitocidal bacteria (Poopathi and Abidha, 20074). More detailed studies include the use of feather peptone for microbial production of carotenoids (Taskin et al. 20115) and protein hydrolysates (Kshetri, 20176). Therefore, the results presented in this work are rather confirmatory, representing a restricted contribution to the field. Additional data should be included to improve this article. In this present form, the manuscript contains preliminary data and lacks sufficient novelty to be considered for indexing.\nIn addition, a limited number or microorganisms was tested. The selection criteria is not clearly described and should be justified. It seems difficult to take solid conclusions since the microbial species tested are limited and only a single strain of each species was tested. Considering that nutritional requirements can be very variable depending on the microorganism, the inclusion of fastidious microorganisms, lactic acid bacteria, and filamentous fungi should be considered.\nSome points on Methods section need to be revised as follows:\nFeathers should be rinsed with abundant water to remove excess detergent before further treatment. Residual detergent could be toxic for some microorganisms. The hydrolysates were air-dried. The reason for not using the oven is unclear since this method was used for blood meal. In this regard, blood meal is mentioned in the Abstract but was not considered in the Discussion. Why was cattle blood used instead of chicken blood? The origin of microbial species tested should be clearly described. Available collection strains should be used in this type of study. The evaluation was performed in liquid medium, supposedly to quantify microbial growth. However, only absorbance at 750 nm (not 600 nm?) was measured, and viable cell counts were not determined. The evaluation of growth on agar plates could give interesting results.\nFinally, the Results section is redundant, since it essentially consists a repetitive description of growth curves. Presentation of results should be carefully reconsidered.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-1918
https://f1000research.com/articles/8-241/v1	01 Mar 19	{ "type": "Method Article", "title": "Introducing high school students to the Gene Ontology classification system", "authors": [ "Mehek Dedhia", "Kenneth Kohetuk", "Wim E. Crusio", "Anna Delprato", "Mehek Dedhia", "Kenneth Kohetuk", "Wim E. Crusio" ], "abstract": "We present an activity that introduces high school students to the Gene Ontology classification system which is widely used in genomics and systems biology studies to characterize large sets of genes based on functional and structural information. This is a valuable and standardized method used to identify genes that act in similar processes and pathways and also to gain insight into the overall architecture and distribution of genes and gene families associated with a particular tissue or disease. Through this exercise, students will learn how the classification system works by analyzing a list of genes using DAVID the Database for Annotation, Visualization and Integrated Discovery that incorporates the Gene Ontology system into its suite of analysis tools. This method of profiling genes is used by our high school student interns to categorize gene expression data related to behavioral neuroscience. Students will get a feel for working with genes and gene sets, gain vocabulary, obtain an understanding of how a database is structured and gain an awareness of the vast amount of information that is known about genes as well as the online analysis tools that are available.", "keywords": [ "gene ontology", "high school students", "genomics" ], "content": "Introduction\n\nGenomics is the branch of biology concerned with the study of genes and their functions (see the National Institutes of Health Frequently Asked Questions about Genetic and Genomic Science). Genomics arose from the acceleration of genetic research which was fueled by the development of rapid and affordable DNA sequencing technologies (Shendure et al., 2017). This opened the door to the sequencing of entire genomes. Presently, the DNA codes for thousands of genomes from diverse species have been sequenced and studied (see the National Center for Biotechnology Information Genome database).\n\nThe goals in genomics research are to address all genes and their inter-relationships in order to understand the combined influence on the function of an organism. With this newfound knowledge of the staggering number of genes that make up an organism, the Gene Ontology (GO) classification system was created to organize genes by their similarities and differences (see Gene Ontology Consortium ‘About’ page). “Ontology” is not a commonly encountered term and there are several definitions that are related to philosophical concepts.\n\nIn the context of information science, as described here, “ontology” is concerned with the representation, formal naming and classification system with the purpose of describing the relationship categories and properties of the data.\n\nThis classification system provides the scientific community with a structured vocabulary for defining genes (Ashburner et al., 2000; du Plessis et al., 2011; Hastings, 2017; Thomas, 2017). GO terms are commonly used in most, if not all, databases and analysis tools relevant to bioinformatics, systems biology (Wanjek, 2011), and genomics studies (du Plessis et al., 2011). GO terms are species specific and are continuously revised and expanded as biological knowledge is obtained (Gaudet et al., 2017).\n\nThe importance of the GO term system becomes apparent when analyzing the organization of genomes and coding regions, the distribution of genes involved in specific processes and the conservation of genes across species (Gaudet et al., 2017). This classification system is also quite powerful when analyzing data from large scale gene expression studies (du Plessis et al., 2011) that consider co-expression data from specific tissues obtained under defined circumstances such as treatment with pharmaceutical agents, or with neurodevelopmental disorders, cancer, or diabetes as examples. GO terms are instrumental for understanding the functions of these genes.\n\nIntroducing GO terms and the gene classification system to high school students will bring them up to speed on a commonly used research tool in current genomics methods and expose them to the vast amounts of data that have been derived from genomics and systems biology studies.\n\nIn the subsequent sections we show an example of how to extract information about a gene from its associated GO terms and then provide instruction for a practical exercise which will enable students to profile a list of genes using GO terms in the bioinformatics resource DAVID, The Database for Annotation, Visualization and Integrated Discovery. This is a protocol that we teach to our high school student interns when they are evaluating gene expression data for their summer projects (Crusio et al., 2017, see BioScience Project student posters).\n\n\nProcedure\n\nThe overall structure of GO is hierarchical and is based on parent-child terms where the parent term is broader and child term is more specialized.\n\nGO terms group genes according to 3 categories, each of which are considered a distinct ontology: Molecular Function (MF, molecular-level activities performed by gene products), Biological Process (BP, the larger processes, or biological programs accomplished by multiple molecular activities), and Cellular Component (CC, the locations relative to cellular structures in which a gene product performs a function).\n\nAs an example, consider the GO term classification for the RAB5A gene (Figure 1). RAB5A belongs to a family of genes called Rab GTPases that are key regulators of intracellular membrane trafficking. Rabs are involved in the formation of transport vesicles and their fusion with membranes. They are enzymes and mediate their function by cycling between a GDP bound inactive and a GTP bound active state. Because of their fundamental and ubiquitous role, this family of genes are associated with many biological processes and diseases.\n\nScreenshot of the DAVID results for the RAB5A gene Top left (blue bar):Gene Symbol identifier. Center (blue bar): full gene name. Labels (left): GO Term BP, GO Term CC, and GO Term MF descriptors. Note that these descriptors are clickable.\n\nThe GO term classification for the RAB5A gene gives:\n\nGOTERM_BP: endocytosis, phagocytosis, small GTPase mediated signal transduction, blood coagulation, protein transport, regulation of endocytosis, synaptic vesicle recycling, viral RNA genome replication, early endosome to late endosome transport, positive regulation of exocytosis, regulation of endosome size, regulation of filopodium assembly, receptor internalization involved in canonical Wnt signaling pathway, regulation of synaptic vesicle exocytosis, regulation of autophagosome assembly\n\nGOTERM_CC: ruffle, intracellular, cytoplasm, endosome, early endosome, cytosol, plasma membrane, synaptic vesicle, endosome membrane, actin cytoskeleton, endocytic vesicle, axon, dendrite, phagocytic vesicle membrane, somatodendritic compartment, melanosome, neuronal cell body, terminal bouton, axon terminus, membrane raft, phagocytic vesicle, extracellular exosome, cytoplasmic side of early endosome membrane.\n\nGOTERM_MF: GTPase activity, protein binding, GTP binding, GDP binding\n\nFrom the RAB5A related GO terms, we get the overall impression that this gene encodes an enzyme, that is involved in signaling, transport and vesicle dynamics and is associated with cell membranes. How do we arrive at this description?\n\nIn this example, the information obtained from the MF category is that the protein product of the RAB5A gene binds to guanine nucleotides: GTP and GDP (Guanosine tri and di phosphate, respectively) and that it is an enzyme. This is evident by the “GTPase activity” term. Whenever the suffix “ase” is used in the context of a gene or protein, it refers to an enzyme, something that catalyzes a chemical reaction. For the BP category, there are several terms associated with intracellular transport, signaling, and endocytosis. Finally, the terms associated with CC include endosome and endosome-like organelles (melanosomes, synaptic vesicles, phagocytic vesicles), as well as membrane structures (ruffles, rafts).\n\nDAVID is primarily a clustering program that groups genes based on different criteria related to GO terms. DAVID links to other databases that contain complementary information like The Gene Ontology. In this exercise, students will use the sample gene lists (DEMOLIST1 or DEMOLIST2) that are accessible from the DAVID database to see how the Gene Ontology classification partitions a set of genes based on GO Terms. Screenshots and videos are provided for step by step instruction. We also provide a video to instruct students on profiling a gene list in DAVID obtained from a random gene list generator.\n\nScreenshot 1 (Figure 2). DAVID landing page. The start analysis link is accessed here and is circled in red in this image. (Video 1 (Delprato et al., 2019a))\n\nScreenshot of the DAVID landing page containing a brief description of the site and links for available tools. The “start analysis” button is circled in red and is located at the top left side of the page. This is the first step in submitting a geneset for profiling with GO Terms in DAVID.\n\nScreenshot 2 (Figure 3). Submitting a gene list. Select either DEMOLIST 1 or DEMOLIST 2 (left panel). The identifier will come up automatically because this is a demonstration list. If you are submitting your own gene list then, the identifier will have to be specified from the dropdown menu (Video 2 (Delprato et al., 2019b)). Typically the identifier is the “Official Gene Symbol”. Click “Gene List”, then “Submit List” (Video 1 (Delprato et al., 2019a)).\n\nScreenshot of the page where a geneset can be submitted “Step1: Enter Gene List”, there are options to copy and paste a geneset or upload a file. There is also an option to use either of two sample lists provided by the DAVID site. Under the pull down menu, “Step 2 Select Identifier”, there are many types of identification designations for the same gene. For the sample lists, the identifier will come up automatically. When submitting a gene list from the geneset generator, the identifier is “Official Gene Symbol” which is used in most cases. If the identifier that you choose does not match the identifiers in the submitted geneset, you will receive a message stating this and an option to convert to the correct identifier.\n\nScreenshot 3 (Figure 4). Species selection. You will see a notice: “Multiple Species, have been Detected”, Highlight “Homo Sapiens” in the window, Select “Homo Sapiens” below the window (Example - DEMOLIST 1: 149 genes, highlighted in grey, left panel). Next, you will see the message “Submission Successful” (Video 1 (Delprato et al., 2019a)).\n\nScreenshot of a successfully submitted geneset. Here it is necessary to select the species. In this instance “Homo sapiens” is highlighted along with the number of genes that are recognized by the site and for which information is available. Once the species is highlighted, it is necessary to click the “Select Species” button in order to limit the output to just the desired species.\n\nScreenshot 4 (Figure 5). Obtaining the results. Select “Functional Annotation Tool”, beneath the blue arrow. Next, select “Functional Annotation Table”, Bottom of the page (Video 1 (Delprato et al., 2019a)).\n\nScreenshot of the different types of analysis options provided by DAVID for a given geneset. For the purpose of this tutorial, the relevant output is the “Functional Annotation Table” located at the bottom of the page.\n\nScreenshot 5 (Figure 6). Reading the output. The gene ID and the full gene name are shown in the blue bars above each entry. The GO Term BP (Biological Process), GO Term CC (Cellular Component), and GO Term MF (Molecular Function), terms are clickable descriptors and link to the Gene Ontology website. See above for a complete description of the GO categories (Video 1 (Delprato et al., 2019a)).\n\nScreenshot of the DAVID “Functional Annotation Table” results for a geneset Top left: (blue bar): Gene Symbol identifier. Center (blue bar): full gene name. Labels (left): GO Term BP, GO Term CC, and GO Term MF descriptors. Note that these descriptors are clickable.\n\nScreenshot 6 (Figure 7). Keyword search. When selecting terms for a keyword search, a more complete outcome is achieved if just a few letters are specified. For example, -”neur” will capture terms both starting with neuro and neural (Video 1 (Delprato et al., 2019a)). DAVID output can be searched for genes related to other process and diseases as well. Have students evaluate the gene list based on their interest. They can identify genes related to a particular process. Students may work individually or in groups.\n\nScreenshot of a keyword search of the DAVID output to identify genes that are relevant to a given biological function. In this case the search term is “neur” (highlighted in yellow), to identify genes related to neurological processes. The keyword search is done via the general search function on your computer. For larger genesets, a computer program may be used.\n\nStudents may wish to try this with their own gene lists. This online gene list generator will enable students to generate a random list of genes for evaluation. (See also Video 2 for instruction (Delprato et al., 2019b))\n\nStep 1. Specify species: Human is the default\n\nStep 2. Specify list length: 200-500 is a good representative number. Note that DAVID will not evaluate lists with more than 2000 genes. An error message stating this will be received.\n\nStep 3. Select “Generate”\n\nStep 4. Copy the gene list using the “Select All” option and paste the list directly into DAVID for evaluation as described above. Make sure to select “Official Gene Symbol” as the identifier when submitting the gene list.\n\nA basic entry and exit ticket method is suggested to determine what students know about genomics and genes before the lesson as well as what they have learned: main points, questions they may have and what they found most interesting. Sample questions are provided in what follows.\n\n1. What is a gene?\n\nA sequence of DNA or RNA which codes for a molecule that has a function. A gene is the basic physical and functional unit of heredity\n\n2. What is genomics?\n\nStudy of the full set of the genes and DNA in an organism\n\n3. How many protein coding genes does a human have?\n\n~20,000\n\n4. Do humans all have the same genes?\n\nYes, but people have different alleles. Alleles are the variation of a gene resulting from mutations. As an example consider eye color. We all have the gene for eye color but some of us have brown, blue or green eyes and there are different shades and hues within those categories.\n\nStep 1. What were the main points of the lesson?\n\nStep 2. Do you have any questions?\n\nStep 3. What aspect of this lesson did you find most interesting?\n\n\nConclusions\n\nWe describe a procedure for students to become acquainted with the Gene Ontology classification systems which is widely used in genomics and systems biology research to characterize gene function. Grouping genes with GO Terms and the DAVID database is based on a protocol that we use with our summer interns to profile gene expression data related to behavioral neuroscience studies (Crusio et al., 2017). Grouping genes in this way identifies genes that function in like processes and also provides information about the overall distribution of a set of genes associated with a particular tissue or process. This tutorial will familiarize early stage students with a biological database and teach them how to mine and extract useful information from a sample list of genes. Entry and exit ticket questions are also included as a formative assessment strategy.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required\n\nExtended data is available from figshare\n\nFigshare: Extended data 1. Video 1: GeneSetProfiling Instructional video for using DAVID to obtain Gene Ontology classifiers for a sample geneset which is provided by the DAVID site https://doi.org/10.6084/m9.figshare.7649225.v1 (Delprato et al., 2019a)\n\nFigshare: Extended data 2. Video 2. UploadGeneSet Instructional video for generating a random geneset and submitting this geneset to DAVID for Gene Ontology classification, https://doi.org/10.6084/m9.figshare.7649231.v1 (Delprato et al., 2019b)", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nAshburner J, Ball CA, Blake JA, et al.: Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet. 2000; 25(1): 25–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCrusio WE, Rubino C, Delprato A: Engaging high school students in neuroscience research -through an e-internship program [version 2; referees: 3 approved]. F1000Res. 2017; 6: 20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDelprato A, Dedhia M, Crusio W, et al.: Video_1 GeneSetProfiling.mp4. figshare. Media. 2019a. http://www.doi.org/10.6084/m9.figshare.7649225.v1\n\nDelprato A, Crusio W, Dedhia M, et al.: Video_2.UploadGeneSist. figshare. Media. 2019b. http://www.doi.org/10.6084/m9.figshare.7649231.v1\n\ndu Plessis L, Skunca N, Dessimoz C: The what, where, how and why of gene ontology--a primer for bioinformaticians. Brief Bioinform. 2011; 12(6): 723–35. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGaudet P, Škunca N, Hu JC, et al.: Primer on the Gene Ontology. In: Dessimoz C, Škunca N. (eds) The Gene Ontology Handbook. Methods Mol Biol. Humana Press, New York, NY. 2017; 1446: 25–37. PubMed Abstract \| Publisher Full Text\n\nHastings J: Primer on Ontologies. In: Dessimoz C, Škunca N. (eds) The Gene Ontology Handbook. Methods Mol Biol. Humana Press, New York, NY. 2017; 1446: 3–13. PubMed Abstract \| Publisher Full Text\n\nShendure J, Balasubramanian S, Church GM, et al.: DNA sequencing at 40: past, present and future. Nature. 2017; 550(7676): 345–353. PubMed Abstract \| Publisher Full Text\n\nThomas PD: The Gene Ontology and the Meaning of Biological Function. In: Dessimoz C, Škunca N. (eds) The Gene Ontology Handbook. Methods Mol Biol. Humana Press, New York, NY. 2017; 1446: 15–24. PubMed Abstract \| Publisher Full Text\n\nWanjek C: Systems Biology as Defined by NIH. The NIH Catalyst. 2011; 19(6). Date accessed Dec 27, 2018. Reference Source" }	[ { "id": "45177", "date": "21 Mar 2019", "name": "William Grisham", "expertise": [ "Reviewer Expertise Neuroscience and pedagogy" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article is aimed at describing an easily searchable database that can be used by students to search by gene ontologies. The point of the article is to provide a guide to use this tool in education at the high school level, but the article could provide a good starting point for college level instructors and even industry trainers. The article describes the underlying question well and chooses a gene, RAB5A, whose function should be readily understood by most biologists.\n\nThe submitted article is timely because the DAVID tool described is not very intuitive. Thus, an easy to follow guide such as the article provides is most welcome. Notably, database tools are becoming more elaborate, and their complexity has created a barrier for effectively use in teaching, particularly at lower levels. The article does a good job of describing this tool, DAVID, in a fashion that will be intelligible to high school and college level students. Lavish illustrations are provided as are links to videos to aid in instruction.\n\nThe use of Entry Ticket and Exit Ticket questions is an excellent pedagogical idea. The questions, however, seem a bit simplistic.\n\nA bigger problem with the article is that there are no data showing how using the DAVID tool in teaching has produced changes in measurable ways. Do students do better on tests of content knowledge? Are students more sophisticated in their thinking as a function of using the DAVID tool? Are they better able to navigate other genomic tools? Do they simply enjoy the lessons more than with traditional instruction? Many of these questions can be answered with a simple pre- and posttest quiz or even comparing current scores on tests/evaluations to historical scores. This addition would make this article of much more value to educators who could then see if incorporating this tool into their pedagogy is worthwhile.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Partly\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? No source data required\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly", "responses": [] }, { "id": "45178", "date": "26 Mar 2019", "name": "Jennifer A. Ufnar", "expertise": [ "Reviewer Expertise Molecular microbiology", "STEM outreach" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article presents an interesting way to introduce gene ontologies and related vocabulary to high school students to enhance biological education. While the proposal describes the method in detail, there is no data presented to show that the method actually works. While this is an interesting protocol, it is not a fully fleshed out methods article. The authors will need to present data to show the viability of the method, or resubmit potentially as a protocol.\n\nA few more specific comments about the article are listed below:\nThe abstract is worded as more of a proposal rather than a paper. The protocol seems very much like following a set of directions (not very thought-provoking or inquiry-based), without relating it to the neuroscience. I think the paper would be much stronger if the authors were to relate it back to the behavioral neuroscience projects that are mentioned in the abstract.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? No\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? No source data required\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? No", "responses": [] } ]	1	https://f1000research.com/articles/8-241
https://f1000research.com/articles/7-1746/v1	05 Nov 18	{ "type": "Systematic Review", "title": "Adverse reactions of dimethyl sulfoxide in humans: a systematic review", "authors": [ "Bennedikte Kollerup Madsen", "Maria Hilscher", "Dennis Zetner", "Jacob Rosenberg", "Maria Hilscher", "Dennis Zetner", "Jacob Rosenberg" ], "abstract": "Background: Dimethyl sulfoxide (DMSO) has been used for medical treatment and as a pharmacological agent in humans since the 1960s. Today, DMSO is used mostly for cryopreservation of stem cells, treatment of interstitial cystitis, and as a penetrating vehicle for various drugs. Many adverse reactions have been described in relation to the use of DMSO, but to our knowledge, no overview of the existing literature has been made. Our aim was to conduct a systematic review describing the adverse reactions observed in humans in relation to the use of DMSO. Methods: This systematic review was reported according to the PRISMA-harms (Preferred Reporting Items for Systematic reviews and Meta-Analysis) guidelines. The primary outcome was any adverse reactions occurring in humans in relation to the use of DMSO. We included all original studies that reported adverse events due to the administration of DMSO, and that had a population of five or more. Results: We included a total of 109 studies. Gastrointestinal and skin reactions were the commonest reported adverse reactions to DMSO. Most reactions were transient without need for intervention. A relationship between the dose of DMSO given and the occurrence of adverse reactions was seen. Conclusions: DMSO may cause a variety of adverse reactions that are mostly transient and mild. The dose of DMSO plays an important role in the occurrence of adverse reactions. DMSO seems to be safe to use in small doses. Registration: PROSPERO CRD42018096117.", "keywords": [ "Dimethyl Sulfoxide", "DMSO", "Adverse reactions", "Toxicology" ], "content": "Introduction\n\nThe first medical report on the use of dimethyl sulfoxide (DMSO) as a pharmacological agent was published in 19641. A year later, the use of DMSO in humans was terminated because experimental studies had shown refractive index changes to the lens of the eye1,2. Years later, DMSO was again approved for use in humans since this side effect was only proven in animal studies2. DMSO has since been used for a variety of purposes, such as treatment of musculoskeletal and dermatological diseases, cryopreservation of stem cells, treatment of interstitial cystitis, treatment of increased intracranial pressure, and many more3–9.\n\nDMSO is a colourless liquid, which is rapidly absorbed when administered dermally or orally10,11. DMSO is used as a cryoprotectant because it decreases osmotic stress and cellular dehydration, and thereby enables stem cells to be stored for several years12. DMSO is mostly excreted through the kidneys, but a small part is excreted through the lungs and liver10. Part of the DMSO is transformed to the volatile metabolite dimethyl sulfide, which gives a characteristic garlic- or oyster-like smell when excreted through the lungs10. DMSO may induce histamine release, which can be the reason for adverse reactions such as flushing, dyspnoea, abdominal cramps, and cardiovascular reactions11.\n\nTo our knowledge, no systematic reviews have been performed on the adverse reactions of DMSO. Our aim was therefore to provide an extensive overview of the possible adverse reactions to DMSO in humans.\n\n\nMethods\n\nOur study-protocol is registered at PROSPERO (Registration number: CRD42018096117). The systematic review was performed according to PRISMA-harms (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines13.\n\nNo limitations were set on the date of publication. The language was restricted to English, Danish, Swedish, Norwegian, and Russian. We included all original studies that administered DMSO to humans and included five or more participants. There was no gender or age restriction. For a study to be included, the authors had to suspect that an observed adverse reaction could be caused by DMSO.\n\nThe primary outcome was any adverse reaction seen in relation to the use of DMSO in humans.\n\nThe search was performed in PubMed (1966-present), EMBASE (1980-present), and the Cochrane Library. The databases were last searched on February 23, 2018. Our search strategy was formulated with the help of a medical research librarian.\n\nThe search string used in PubMed was: ((dimethyl sulfoxide) OR DMSO) AND ((((((administration and dosage) OR adverse reactions) OR alternate effects) OR secondary response) OR toxicology) OR side effects)). The search was restricted to humans. The search string was adapted to EMBASE and Cochrane Library using the same search-words as abovementioned.\n\nThe search string used in EMBASE was: ((dmso or dimethyl sulfoxide) and ((side effect or toxicology or secondary response or alternate effects or alternate reactions or (administration and dosage)) and (dmso or dimethyl sulfoxide))).mp. The search was restricted to humans, articles and medline journals were excluded.\n\nThe search string used in Cochrane was: (adverse drug events and dimethyl sulfoxide). The search was restricted to trials.\n\nTwo authors (B.K.M. and D.Z.) independently screened title and abstract according to the eligibility criteria using www.covidence.org. Discrepancies were resolved by discussion. One author screened the full-text articles (B.K.M.). Russian articles were screened by an author fluent in Russian (M.H.).\n\nAfter the screening process was finished, all included studies were imported to an Excel sheet (Microsoft Excel 2016). Data extraction was performed by two authors. Data extracted were: author, publication year, country, study characteristics (study design, sample size, size of comparison group if present, time to follow-up), use of DMSO (reason for use, treatment duration, administration route, dose of DMSO), and adverse reactions observed (number of persons experiencing an adverse reaction, method of assessing, and duration of adverse reaction).\n\nThe Newcastle-Ottawa-Scale was used to assess the risk of bias in non-randomized observational studies14. Risk of bias in randomized controlled trials was assessed using the Cochrane Handbook “Risk of Bias” assessment tool15. Risk of bias was assessed at the outcome level.\n\nThe primary summary measure was percentage of persons experiencing an adverse reaction, as well as the range in which a reaction occurred in the studies included. No meta-analysis and further summery measures were planned due to the expected large heterogeneity of the studies.\n\n\nResults\n\nOur primary search identified 2599 studies (Figure 1). After the evaluation process, 109 studies were included in the final review2,4,6–9,16–118.\n\nGastrointestinal adverse reactions, possibly due to DMSO, were reported in 61 studies. Of these, 10 studies were randomized controlled trials16,30,33,55,57,59,67,79,93,95, 49 were cohort studies2,4,7,9,18,19,23,25–27,29,35,38–43,45,46,48,50–54,58,60,66,68,69,71,73,83,85–88,90,94,97,98,101,104,105,112,113,115,118, and 2 were case series84,109. Most studies reported the number of patients experiencing an adverse reaction (Table 1). Other studies reported adverse reactions observed in relation to the number of treatments given (Table 2).\n\n†Incidences of the adverse reactions have been calculated for all the individual studies. (min%–max%) are the lowest and highest observed incidence of an adverse reaction observed in the group of studies included.\n\n‡ Nausea and vomiting are reported as one combined adverse reaction in some studies.\n\n† Incidences of the adverse reactions have been calculated for all the individual studies. (min%–max%) are the lowest and highest observed incidence of an adverse reaction.\n\n‡ Intravenous administration.\n\nThe most commonly reported gastrointestinal adverse reactions were nausea and vomiting. The incidence of nausea seems to be less common with the transdermal administration of DMSO compared with intravenous administration. The majority of studies reported an incidence of nausea between 2–14%, with the exception of one study, reporting an incidence of 32%2. In one study that failed to specify the dose, 8 of 42 patients reported nausea and anorexia, but the symptoms disappeared in five of the eight patients when the dose of DMSO was reduced45.\n\nOften the studies had short follow-up periods (less than 24 hours), especially when DMSO was used as a cryoprotectant. The study reporting the highest incidence of nausea had a follow-up period of 5 days48, and the authors concluded that the high incidence of nausea observed might be due to the long follow-up period48. In another article using the same data119, it was suggested that the delayed nausea was due to gastrointestinal mucosal damage, and only the initial nausea could be related to DMSO, and therefore we decided only to include the data from the first 2 days after infusion48.\n\nHalitosis was reported in 29 studies4,9,16,19,29,30,35,42,43,45,50,52,55,57,58,66–68,79,83,85,88,94,95,97,98,109,112,113. In some studies, patients discontinued treatment due to halitosis9,45,83,94. In five studies, all patients experienced halitosis9,45,83,94. Unlike halitosis, other gastrointestinal side effects were reported more often when DMSO was administered intravenously, than transdermally or intravesically.\n\nOne study reported a severe case of nausea, vomiting, and abdominal cramps in one patient with an acute allergic reaction59. However, in most studies the reported gastrointestinal reactions were transient and mild, often lasting only minutes to a couple of hours16,38,41,68,85,87,90. Several studies reported a relationship between the dose of DMSO and the occurrence of gastrointestinal adverse reactions26,33,53,73,83,85.\n\nCardiovascular and respiratory adverse reactions were reported in 33 studies. Of these, two were randomized controlled trials33,59, 30 were cohort studies7,18,23,25–27,36,39–41,51,54,61,65,66,68,73,74,80,85–87,90,100–102,104,115,117, and one was a preliminary report91. Except for one study66, all studies reporting cardiovascular and respiratory reactions administered DMSO intravenously (Table 3 and Table 4).\n\n†Incidences of the adverse reactions have been calculated for all the individual studies. (min%–max%) are the lowest and highest observed incidence of an adverse reaction.\n\n‡ DMSO was administered intravenously in all studies.\n\n§ DMSO was administered intravenously in all studies. Horacek et al. [58] measured 42 patients with an increase in systolic blood pressure, and 31 patients with an increase in diastolic blood pressure. This was counted as 73 cases of hypertension.\n\n¶ In both studies, asystole occurred because of DMSO effect on the trigeminal nerve and activation of the trigeminal cardiac reflex. d) in one study DMSO was administered transdermally\n\n†Incidences of the adverse reactions have been calculated for all the individual studies. (%, min-max) are the lowest and highest observed incidence of an adverse reaction observed in the group of studies included.\n\n‡ DMSO was administered intravenously.\n\nBradycardia was defined as a heart rate less than 60 beats per minute41,61 and was often transient23,61,90,115, but cases where atropine was needed are described49,96. A lowered heart rate not enough to be considered bradycardia was observed in four studies39,41,54,61.\n\nIn some studies, hypertension did not require intervention61,102, but cases where medication was needed to control the hypertension, or where treatment was stopped due to hypertension, are described41,54,85. Hypotension was also described as transient most of the time18,23,68,87,104, with some cases needing intervention40,51,54.\n\nOne study reported 11 cases of transient extrasystoles in 22 patients receiving cryopreserved autologous blood stem cells, monitored with Holter during infusion73. There were two studies reporting cases of asystole during embolization of dural arteriovenous fistulas with a substance called Onyx, a non-adhesive liquid embolic agent dissolved in DMSO91,100.\n\nDyspnea was reported in seven studies18,25,27,54,66,68,85. A single study reported eight patients with transient shock after stem cell transfusion51. Some of these patients developed loss of consciousness and cyanosis but recovered promptly and had no need for additional therapy, whereas the rest of the patients developed severe hypotension or transient dyspnea, which was described as the reason for the transient shock. Further description of the condition was not provided.\n\nSeveral of the studies found a correlation between the dose of DMSO used and the incidence of cardiovascular adverse reactions41,67,71,75,78,85,86,93,101,115.\n\nDermatological side effects are common when DMSO is administered transdermally. Skin reactions or allergic reactions were reported in 58 studies. DMSO was applied transdermally in 43 studies2,4,6,17,19–22,24,28–32,37,44,45,52,55,57,63,64,66,67,69,72,75,76,78,79,82,83,88,89,93,95,96,106,108,109,111–113, intravenously in 14 studies25,40,41,51,59,73,74,77,85,86,92,98,101,110 and intraarticular in one103 (Table 5).\n\n†Incidences of the adverse reactions have been calculated for all the individual studies. (min%–max%) are the lowest and highest observed incidence of an adverse reaction observed in the group of studies included.\n\n‡ Transdermal application only.\n\n§ One study administered DMSO through intraarticular injection [38].\n\n¶ DMSO was administered intravenously in all studies.\n\nThe most common skin reaction was a local burning sensation reported in 13 studies17,21,24,28,30,45,55,57,67,69,79,93,106. In one study, all participants experienced this burning sensation45. In the same study, four participants experienced a transient peripheral edema associated with itching and erythema45. A single study described a burning sensation in four of 669 patients when DMSO was given as a local injection92; another study described burning in two out of 17 patients when DMSO was injected intraarticularly103.\n\nMost skin reactions were transient, only lasting minutes17,24,32,67,72, but some studies reported cases described as serious, causing discontinuation of treatment2,6,52,63,78,96. There were two studies describing that skin reactions to DMSO would disappear after days of continuous treatment45,83. Another study reported that 1 of 18 patients treated for psoriasis with DMSO was hospitalized due to exfoliative erythroderma63. In another study, two patients, diagnosed with dermographia developed prominent areas of weals after DMSO application95.\n\nAcute allergic reactions due to use of DMSO were reported in six studies37,44,59,86,98,110. One study reported that 63 of 144 patients experienced allergic reactions, which was not described as serious adverse events (bronchospasms, facial flushing, rash)86. In two other studies, acute allergic reactions were characterized as serious adverse events59,110.\n\nFlushing was regarded as an allergic reaction in this review and was only reported when DMSO was administered intravenously25,40,41,51,54,73,74. A total of four studies, not depicted in Table 5, reported 204 cases of flushing during 1439 stem cell infusions25,40,51,74. Several studies observed a relationship between the dose of DMSO and the occurrence of adverse reactions67,75,78,83,88,93.\n\nHeadache is the most common neurological adverse reaction reported. In one study, headache was the reason for withdrawal of 2 out of 21 patients being treated with DMSO116.\n\nThree studies using DMSO as a cryoprotectant in stem cell transfusions described seizures after administration18,36,47. Severe encephalopathy was observed in one patient99, and transient cranial nerve III and IV palsy was observed in one patient after Onyx embolization34. One study described neurological symptoms occurring during and after transfusion, but they did not define neurological symptoms in detail86.\n\nFew urogenital reactions were described (Table 6 and Table 7). Hemoglobinuria was described as an adverse reaction seen after transfusion of stem cell products39,51,56,73. However, hemoglobinuria is often attributed to erythrocyte debris in the transplant material and has thus not been interpreted as being caused by DMSO39,73. The other urogenital reactions (Table 6 and Table 7) all occurred after DMSO instillation in the bladder38,49,97.\n\n†Incidences of the adverse reactions have been calculated for all the individual studies. (min%–max%) are the lowest and highest observed incidence of an adverse reaction observed in the group of studies included.\n\n†Incidences of the adverse reactions have been calculated for all the individual studies. (%, min-max) are the lowest and highest observed incidence of an adverse reaction observed in the group of studies included.\n\nOnly one study in this review administered DMSO as eye-drops114. In this study, two patients experienced severe conjunctival hyperemia due to allergic reactions, and 25% of patients experienced a stinging sensation when eye-drops were applied114. Other studies performed eye examinations to determine whether DMSO caused changes in the lens; however, no such cases were observed2,45.\n\nHyponatremia occurred in six patients after they received large doses of DMSO as treatment for cranial hypertension62. This adverse reaction was not reported in other studies (Table 8).\n\n†Incidences of the adverse reactions have been calculated for all the individual studies. (%, min-max) are the lowest and highest observed incidence of an adverse reaction observed in the group of studies included.\n\nVery few cases of serious adverse reactions associated with DMSO have been described18,36,51,59.\n\nIn this review, we included 76 cohort studies, of which 64 were prospective2,4,6,7,20,22,24–27,29,31,32,34–38,40–45,48,50–54,56,58,60,63,65,66,68–70,72,73,77,80,81,83,85,88,90,92,94,97,98,101–104,107,108,110,112,115,117,118 and 13 were retrospective9,18,23,39,46,47,61,71,74,86,87,100,105. Bias was assessed using The Newcastle-Ottawa-Scale14. Using this scale, studies were given zero to nine stars. A high number of stars equals low risk of bias and vice versa. The studies in this review had a median value of 5 stars, with a range of 2–8. No studies received the highest possible value of nine stars. Very few studies had a comparison group that did not receive DMSO, and often the occurrence of adverse reactions was poorly described. There were 24 randomized controlled trials (Figure 2). Many studies received an unclear risk of bias because often it was vaguely described how adverse reactions were reported.\n\nOverall, there was a high risk of bias when assessing the description of adverse reactions. Some studies were not assessed for bias due to being case-reports, preliminary trials, or because they included more than one study design17,19,28,62,84,91,99,109,111,113.\n\n\nDiscussion\n\nGastrointestinal and dermatological adverse reactions were the most commonly reported in the included studies. Cardiac adverse reactions only occurred when DMSO was administered intravenously, whereas dermatological reactions mostly occurred when DMSO was administered on the skin. Serious neurological and cardiac reactions were rare and only described in few studies. There seems to be a dose-response relationship between DMSO and adverse reactions with no or mild reactions in low doses.\n\nMany studies on the use of DMSO have been performed in Russia. These studies have not been readily accessible to the global community due to the language barrier. In this review, we have included not only studies dating back almost 50 years, but also articles written in Russian, which is an important strength of the review. Some studies used the NCI-CTC (National Cancer Institute’s Common Terminology Criteria for adverse events), but often no scale was used, and the occurrence of adverse reactions were poorly reported. It was difficult to make conclusions on the frequency of a specific adverse reaction, because the exact number of patients experiencing a reaction was often not stated. Also, several studies using DMSO as a cryoprotectant concluded that other factors affected the occurrence of adverse reactions7,85,86. One study prospectively looked at the adverse reactions observed in relation to autologous transplantation in 64 European Blood and Marrow Transplant Group centers7. They had difficulties isolating the effects of DMSO from confounding factors such as cell breakdown products and conditioning chemotherapy. Factors such as age, gender, volume transfused, granulocyte concentration, clumping of transplant material, and amount of red blood cells played a role in the occurrence of adverse reactions61,86,120–122. Another study believed that acute volume expansion, electrolyte imbalance and vagal responses to the coldness of the freshly thawed infusate were more likely reasons for cardiac arrhythmias during stem cell transfusions than the DMSO infused123. This differs from other studies, which found a clear connection between dose of DMSO and occurrence of cardiac adverse reactions41,67,71,75,78,85,86,93,101,115. Therefore, it is possible that some adverse reactions are more or less common than found in this review. The rarer side effects are often reported in case reports, which often did not meet the eligibility criteria in this review. However, we have included several larger studies in this review, and they found a very small occurrence of serious adverse events7,55,66,74.\n\nIn conclusion, adverse reactions due to DMSO are often mild and transient and do not qualify as serious adverse events. Cardiovascular and respiratory adverse reactions occur mostly when DMSO is administered intravenously, whereas dermatological reactions have a higher incidence when DMSO is administered transdermally. An important finding is that the occurrence of adverse reactions seems to be related to the dose of DMSO, and it therefore seems safe to continue the use of DMSO in small doses.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nSupplementary material\n\nSupplementary File 1. Completed PRISMA harms checklist.\n\nClick here to access the data\n\n\nReferences\n\nBrown JH: Dimethyl sulfoxide (DMSO)--a unique therapeutic entity. Aviat Space Environ Med. 1982; 53(1): 82–8. PubMed Abstract\n\nBrobyn RD: The human toxicology of dimethyl sulfoxide. Ann N Y Acad Sci. 1975; 243: 497–506. PubMed Abstract \| Publisher Full Text\n\nSwanson BN: Medical use of dimethyl sulfoxide (DMSO). Rev Clin Basic Pharm. 1985; 5(1–2): 1–33. PubMed Abstract\n\nPaul MM: Interval therapy with dimethyl sulfoxide. Ann N Y Acad Sci. 1967; 141(1): 586–98. PubMed Abstract \| Publisher Full Text\n\nKulah A, Akar M, Baykut L: Dimethyl sulfoxide in the management of patient with brain swelling and increased intracranial pressure after severe closed head injury. Neurochirurgia (Stuttg). 1990; 33(6): 177–80. PubMed Abstract \| Publisher Full Text\n\nRosenbaum EE, Herschler RJ, Jacob SW: Dimethyl Sulfoxide in Musculoskeletal Disorders. JAMA. 1965; 192: 309–13. PubMed Abstract \| Publisher Full Text\n\nMorris C, de Wreede L, Scholten M, et al.: Should the standard dimethyl sulfoxide concentration be reduced? Results of a European Group for Blood and Marrow Transplantation prospective noninterventional study on usage and side effects of dimethyl sulfoxide. Transfusion. 2014; 54(10): 2514–22. PubMed Abstract \| Publisher Full Text\n\nPeeker R, Haghsheno MA, Holmäng S, et al.: Intravesical bacillus Calmette-Guerin and dimethyl sulfoxide for treatment of classic and nonulcer interstitial cystitis: a prospective, randomized double-blind study. J Urol. 2000; 164(6): 1912–5, discussion 1915–6. PubMed Abstract \| Publisher Full Text\n\nAmemori S, Iwakiri R, Endo H, et al.: Oral dimethyl sulfoxide for systemic amyloid A amyloidosis complication in chronic inflammatory disease: a retrospective patient chart review. J Gastroenterol. 2006; 41(5): 444–9. PubMed Abstract \| Publisher Full Text\n\nHucker HB, Miller JK, Hochberg A, et al.: Studies on the absorption, excretion and metabolism of dimethylsulfoxide (DMSO) in man. J Pharmacol Exp Ther. 1967; 155(2): 309–17. PubMed Abstract\n\nKligman AM: Topical Pharmacology and Toxicology of Dimethyl Sulfoxide. 1. JAMA. 1965; 193: 796–804. PubMed Abstract \| Publisher Full Text\n\nLovelock JE, Bishop MW: Prevention of freezing damage to living cells by dimethyl sulphoxide. Nature. 1959; 183(4672): 1394–5. PubMed Abstract \| Publisher Full Text\n\nZorzela L, Loke YK, Ioannidis JP, et al.: PRISMA harms checklist: improving harms reporting in systematic reviews. BMJ. 2016; 352: i157. PubMed Abstract \| Publisher Full Text\n\nWells G, Shea B, O’Connell D, et al.: The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. Reference Source\n\nHiggins JPT, Altman DG, Sterne JAC: Chapter 8: Assessing risk of bias in included studies. In: Higgins JPT GS. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration. 2008. Publisher Full Text\n\nVadhan-Raj S, Kavanagh JJ, Freedman RS, et al.: Safety and efficacy of transfusions of autologous cryopreserved platelets derived from recombinant human thrombopoietin to support chemotherapy-associated severe thrombocytopenia: a randomised cross-over study. Lancet. 2002; 359(9324): 2145–52. PubMed Abstract \| Publisher Full Text\n\nOgden HD: Experiences with DMSO in treatment of headache. Ann N Y Acad Sci. 1967; 141(1): 646–8. PubMed Abstract \| Publisher Full Text\n\nMartin-Henao GA, Resano PM, Villegas JM, et al.: Adverse reactions during transfusion of thawed haematopoietic progenitor cells from apheresis are closely related to the number of granulocyte cells in the leukapheresis product. Vox Sang. 2010; 99(3): 267–73. PubMed Abstract \| Publisher Full Text\n\nSavastano AA: Clinical experiences with dimethyl sulfoxide (DMSO) in human subjects. Approval must be withheld until safety in extended use is established. R I Med J. 1984; 67(3): 119–21. PubMed Abstract\n\nIvanov OL, Potekaev NS, Aliab’eva AP: [Dimethyl sulfoxide applications in the treatment of erythema nodosum]. Ter Arkh. 1983; 55(9): 104–7. PubMed Abstract\n\nVuopala U, Isomäki H, Kaipainen WJ: Dimethyl sulfoxide (DMSO) ointment in the treatment of rheumatoid arthritis. A double blind study. Acta Rheumatol Scand. 1969; 15(2): 139–44. PubMed Abstract \| Publisher Full Text\n\nMenne T: Nickel allergy--reliability of patch test. Evaluated in female twins. Derm Beruf Umwelt. 1981; 29(6): 156–60. PubMed Abstract\n\nCastillo N, Garcia-Cadenas I, Garcia O, et al.: Few and nonsevere adverse infusion events using an automated method for diluting and washing before unrelated single cord blood transplantation. Biol Blood Marrow Transplant. 2015; 21(4): 682–7. PubMed Abstract \| Publisher Full Text\n\nPandhi R, Kaur I, Kumar B: Lack of effect of dimethylsulphoxide in cutaneous amyloidosis. J Dermatolog Treat. 2002; 13(1): 11–4. PubMed Abstract \| Publisher Full Text\n\nHalle P, Tournilhac O, Knopinska-Posluszny W, et al.: Uncontrolled-rate freezing and storage at -80 degrees C, with only 3.5-percent DMSO in cryoprotective solution for 109 autologous peripheral blood progenitor cell transplantations. Transfusion. 2001; 41(5): 667–73. PubMed Abstract \| Publisher Full Text\n\nSyme R, Bewick M, Stewart D, et al.: The role of depletion of dimethyl sulfoxide before autografting: on hematologic recovery, side effects, and toxicity. Biol Blood Marrow Transplant. 2004; 10(2): 135–41. PubMed Abstract \| Publisher Full Text\n\nOzdemir E, Akgedik K, Akdogan S, et al.: The lollipop with strawberry aroma may be promising in reduction of infusion-related nausea and vomiting during the infusion of cryopreserved peripheral blood stem cells. Biol Blood Marrow Transplant. 2008; 14(12): 1425–8. PubMed Abstract \| Publisher Full Text\n\nSilvestri DL, Corey L, Holmes KK: Ineffectiveness of topical idoxuridine in dimethyl sulfoxide for therapy for genital herpes. JAMA. 1982; 248(8): 953–9. PubMed Abstract \| Publisher Full Text\n\nShainhouse JZ, Grierson LM, Naseer Z: A long-term, open-label study to confirm the safety of topical diclofenac solution containing dimethyl sulfoxide in the treatment of the osteoarthritic knee. Am J Ther. 2010; 17(6): 566–76. PubMed Abstract \| Publisher Full Text\n\nSimon LS, Grierson LM, Naseer Z, et al.: Efficacy and safety of topical diclofenac containing dimethyl sulfoxide (DMSO) compared with those of topical placebo, DMSO vehicle and oral diclofenac for knee osteoarthritis. Pain. 2009; 143(3): 238–45. PubMed Abstract \| Publisher Full Text\n\nEngel MF: Dimethyl sulfoxide in the treatment of scleroderma. South Med J. 1972; 65(1): 71–3. PubMed Abstract\n\nLudwig CU, Stoll HR, Obrist R, et al.: Prevention of cytotoxic drug induced skin ulcers with dimethyl sulfoxide (DMSO) and alpha-tocopherole. Eur J Cancer Clin Oncol. 1987; 23(3): 327–9. PubMed Abstract \| Publisher Full Text\n\nMitrus I, Smagur A, Fidyk W, et al.: Reduction of DMSO concentration in cryopreservation mixture from 10% to 7.5% and 5% has no impact on engraftment after autologous peripheral blood stem cell transplantation: results of a prospective, randomized study. Bone Marrow Transplant. 2018; 53(3): 274–280. PubMed Abstract \| Publisher Full Text\n\nElhammady MS, Wolfe SQ, Farhat H, et al.: Onyx embolization of carotid-cavernous fistulas. J Neurosurg. 2010; 112(3): 589–94. PubMed Abstract \| Publisher Full Text\n\nHung MJ, Chen YT, Shen PS, et al.: Risk factors that affect the treatment of interstitial cystitis using intravesical therapy with a dimethyl sulfoxide cocktail. Int Urogynecol J. 2012; 23(11): 1533–9. PubMed Abstract \| Publisher Full Text\n\nLemarie C, Camels B, Malenfant C, et al.: Clinical experience with the delivery of thawed and washed autologous blood cells, with an automated closed fluid management device: CytoMate. Transfusion. 2005; 45(5): 737–42. PubMed Abstract \| Publisher Full Text\n\nMendelow AY, Forsyth A, Florence AT, et al.: Patch testing for nickel allergy. The influence of the vehicle on the response rate to topical nickel sulphate. Contact Dermatitis. 1985; 13(1): 29–33. PubMed Abstract \| Publisher Full Text\n\nFowler JE Jr: Prospective study of intravesical dimethyl sulfoxide in treatment of suspected early interstitial cystitis. Urology. 1981; 18(1): 21–6. PubMed Abstract \| Publisher Full Text\n\nPerseghin P, Balduzzi A, Bonanomi S, et al.: Infusion-related side-effects in children undergoing autologous hematopoietic stem cell transplantation for acute leukemia. Bone Marrow Transplant. England; 2000; 26(1): 116–8. PubMed Abstract \| Publisher Full Text\n\nRowley SD, Feng Z, Yadock D, et al.: Post-thaw removal of DMSO does not completely abrogate infusional toxicity or the need for pre-infusion histamine blockade. Cytotherapy. 1999; 1(6): 439–46. PubMed Abstract \| Publisher Full Text\n\nDavis J, Rowley SD, Santos GW: Toxicity of autologous bone marrow graft infusion. Prog Clin Biol Res. 1990; 333: 531–40. PubMed Abstract\n\nBertelli G, Gozza A, Forno GB, et al.: Topical dimethylsulfoxide for the prevention of soft tissue injury after extravasation of vesicant cytotoxic drugs: a prospective clinical study. J Clin Oncol. 1995; 13(11): 2851–5. PubMed Abstract \| Publisher Full Text\n\nBarker SB, Matthews PN, Philip PF, et al.: Prospective study of intravesical dimethyl sulphoxide in the treatment of chronic inflammatory bladder disease. Br J Urol. 1987; 59(2): 142–4. PubMed Abstract \| Publisher Full Text\n\nTseraidis GS, Popova SS, Zykova NIa: [Use of a progesterone ointment in hirsutism]. Vestn Dermatol Venerol. 1986; (7): 14–7. PubMed Abstract\n\nScherbel AL, McCormack LJ, Layle JK: Further observations on the effect of dimethyl sulfoxide in patients with generalized scleroderma. (Progressive systemic sclerosis). Ann N Y Acad Sci. 1967; 141(1): 613–29. PubMed Abstract \| Publisher Full Text\n\nMitrus I, Smagur A, Giebel S, et al.: A faster reconstitution of hematopoiesis after autologous transplantation of hematopoietic cells cryopreserved in 7.5% dimethyl sulfoxide if compared to 10% dimethyl sulfoxide containing medium. Cryobiology. 2013; 67(3): 327–31. PubMed Abstract \| Publisher Full Text\n\nMueller LP, Theurich S, Christopeit M, et al.: Neurotoxicity upon infusion of dimethylsulfoxide-cryopreserved peripheral blood stem cells in patients with and without pre-existing cerebral disease. Eur J Haematol. 2007; 78(6): 527–31. PubMed Abstract \| Publisher Full Text\n\nGonella S, Di Giulio P: Delayed chemotherapy-induced nausea and vomiting in autologous hematopoietic cell transplant patients: an exploratory analysis. Tumori. 2015; 101(6): e154–9. PubMed Abstract \| Publisher Full Text\n\nCervigni M, Sommariva M, Tenaglia R, et al.: A randomized, open-label, multicenter study of the efficacy and safety of intravesical hyaluronic acid and chondroitin sulfate versus dimethyl sulfoxide in women with bladder pain syndrome/interstitial cystitis. Neurourol Urodyn. 2017; 36(4): 1178–86. PubMed Abstract \| Publisher Full Text\n\nVuopala U, Kaipainen WJ: DMOS in the treatment of Dupuytren’s contracture. A therapeutic experiment. Acta Rheumatol Scand. 1971; 17(1): 61–2. PubMed Abstract \| Publisher Full Text\n\nOkamoto Y, Takaue Y, Saito S, et al.: Toxicities associated with cryopreserved and thawed peripheral blood stem cell autografts in children with active cancer. Transfusion. 1993; 33(7): 578–81. PubMed Abstract \| Publisher Full Text\n\nZuckner J, Uddin J, Gantner GE Jr: Local application of dimethyl sulfoxide and DMSO combined with triamcinolone acetonide in rheumatoid arthritis. Ann N Y Acad Sci. 1967; 141(1): 555–9. PubMed Abstract \| Publisher Full Text\n\nAkkök CA, Liseth K, Nesthus I, et al.: Autologous peripheral blood progenitor cells cryopreserved with 5 and 10 percent dimethyl sulfoxide alone give comparable hematopoietic reconstitution after transplantation. Transfusion. 2008; 48(5): 877–83. PubMed Abstract \| Publisher Full Text\n\nDavis JM, Rowley SD, Braine HG, et al.: Clinical toxicity of cryopreserved bone marrow graft infusion. Blood. 1990; 75(3): 781–6. PubMed Abstract\n\nRoth SH, Shainhouse JZ: Efficacy and safety of a topical diclofenac solution (pennsaid) in the treatment of primary osteoarthritis of the knee: a randomized, double-blind, vehicle-controlled clinical trial. Arch Intern Med. 2004; 164(18): 2017–23. PubMed Abstract \| Publisher Full Text\n\nGalmés A, Besalduch J, Bargay J, et al.: A simplified method for cryopreservation of hematopoietic stem cells with -80 degrees C mechanical freezer with dimethyl sulfoxide as the sole cryoprotectant. Leuk Lymphoma. 1995; 17(1–2): 181–4. PubMed Abstract \| Publisher Full Text\n\nBookman AA, Williams KS, Shainhouse JZ: Effect of a topical diclofenac solution for relieving symptoms of primary osteoarthritis of the knee: a randomized controlled trial. CMAJ. 2004; 171(4): 333–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAliab’eva AP, Melikhova NI, Murav’ev IuV: [Use of heparin applications in a dimethyl sulfoxide medium in the overall treatment of rheumatoid arthritis in children]. Pediatriia. 1980; (9): 50–1. PubMed Abstract\n\nShpall EJ, LeMaistre CF, Holland K, et al.: A prospective randomized trial of buffy coat versus CD34-selected autologous bone marrow support in high-risk breast cancer patients receiving high-dose chemotherapy. Blood. 1997; 90(11): 4313–20. PubMed Abstract\n\nFuks JZ, Egorin MJ, Aisner J, et al.: Cyclophosphamide and dimethylsulfoxide in the treatment of squamous carcinoma of the lung. Therapeutic efficacy, toxicity, and pharmacokinetics. Cancer Chemother Pharmacol. 1981; 6(2): 117–20. PubMed Abstract \| Publisher Full Text\n\nStyler MJ, Topolsky DL, Crilley PA, et al.: Transient high grade heart block following autologous bone marrow infusion. Bone Marrow Transplant. 1992; 10(5): 435–8. PubMed Abstract\n\nMarshall LF, Camp PE, Bowers SA: Dimethyl sulfoxide for the treatment of intracranial hypertension: a preliminary trial. Neurosurgery. 1984; 14(6): 659–63. PubMed Abstract \| Publisher Full Text\n\nEngel MF: Dimethyl sulfoxide (DMSO) in clinical dermatology. South Med J. 1966; 59(11): 1318–9. PubMed Abstract\n\nRoth SH, Fuller P: Pooled safety analysis of diclofenac sodium topical solution 1.5% (w/w) in the treatment of osteoarthritis in patients aged 75 years or older. Clin Interv Aging. 2012; 7: 127–37. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLv X, Jiang C, Li Y, et al.: Percutaneous transvenous packing of cavernous sinus with Onyx for cavernous dural arteriovenous fistula. Eur J Radiol. 2009; 71(2): 356–62. PubMed Abstract \| Publisher Full Text\n\nDemos CH, Beckloff GL, Donin MN, et al.: Dimethyl sulfoxide in musculoskeletal disorders. Ann N Y Acad Sci. 1967; 141(1): 517–23. PubMed Abstract \| Publisher Full Text\n\nSpruance SL, Stewart JC, Freeman DJ, et al.: Early application of topical 15% idoxuridine in dimethyl sulfoxide shortens the course of herpes simplex labialis: a multicenter placebo-controlled trial. J Infect Dis. 1990; 161(2): 191–7. PubMed Abstract \| Publisher Full Text\n\nBojanic I, Cepulic BG, Mazic S, et al.: Toxicity related to autologous peripheral blood haematopoietic progenitor cell infusion is associated with number of granulocytes in graft, gender and diagnosis of multiple myeloma. Vox Sang. 2008; 95(1): 70–5. PubMed Abstract \| Publisher Full Text\n\nBlumenthal LS, Fuchs M: The clinical use of dimethyl sulfoxide on various headaches, musculoskeletal, and other general medical disorders. Ann N Y Acad Sci. 1967; 141(1): 572–85. PubMed Abstract \| Publisher Full Text\n\nHoang BX, Le BT, Tran HD, et al.: Dimethyl sulfoxide-sodium bicarbonate infusion for palliative care and pain relief in patients with metastatic prostate cancer. J Pain Palliat Care Pharmacother. 2011; 25(4): 350–5. PubMed Abstract \| Publisher Full Text\n\nStroncek DF, Fautsch SK, Lasky LC, et al.: Adverse reactions in patients transfused with cryopreserved marrow. Transfusion. 1991; 31(6): 521–6. PubMed Abstract \| Publisher Full Text\n\nKaidbey KH: Therapy of resistant psoriasis with topical corticosteroids and dimethylsulfoxide. Dermatologica. 1976; 152(5): 316–20. PubMed Abstract \| Publisher Full Text\n\nZambelli A, Poggi G, Da Prada G, et al.: Clinical toxicity of cryopreserved circulating progenitor cells infusion. Anticancer Res. 1998; 18(6B): 4705–8. PubMed Abstract\n\nOtrock ZK, Sempek DS, Carey S, et al.: Adverse events of cryopreserved hematopoietic stem cell infusions in adults: a single-center observational study. Transfusion. 2017; 57(6): 1522–6. PubMed Abstract \| Publisher Full Text\n\nMatsumoto J: Clinical trials of dimethyl sulfoxide in rheumatoid arthritis patients in Japan. Ann N Y Acad Sci. 1967; 141(1): 560–8. PubMed Abstract \| Publisher Full Text\n\nMelikhova NI, Murav’ev IuV, Sigidin IaA, et al.: [Effectiveness of the treatment of juvenile rheumatoid arthritis with dimethyl sulfoxide gel]. Pediatriia. 1986; (6): 53–4. PubMed Abstract\n\nHolbro A, Graf L, Topalidou M, et al.: Cryopreserved stem cell products containing dimethyl sulfoxide lead to activation of the coagulation system without any impact on engraftment. Transfusion. 2014; 54(6): 1508–14. PubMed Abstract \| Publisher Full Text\n\nWilliams HJ, Furst DE, Dahl SL, et al.: Double-blind, multicenter controlled trial comparing topical dimethyl sulfoxide and normal saline for treatment of hand ulcers in patients with systemic sclerosis. Arthritis Rheum. 1985; 28(3): 308–14. PubMed Abstract \| Publisher Full Text\n\nSimpson JR: Idoxuridine in the treatment of herpes zoster. Practitioner. 1975; 215(1286): 226–9. PubMed Abstract\n\nLv X, Li Y, Jiang C, et al.: The incidence of trigeminocardiac reflex in endovascular treatment of dural arteriovenous fistula with onyx. Interv Neuroradiol. 2010; 16(1): 59–63. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHoang BX, Tran DM, Tran HQ, et al.: Dimethyl sulfoxide and sodium bicarbonate in the treatment of refractory cancer pain. J Pain Palliat Care Pharmacother. 2011; 25(1): 19–24. PubMed Abstract \| Publisher Full Text\n\nBosso JA, Spruance SL, Wenerstrom G: Tolerance and percutaneous absorption of topically applied arildone. J Clin Pharmacol. 1985; 25(2): 95–9. PubMed Abstract \| Publisher Full Text\n\nOzkaya-Bayazit E, Kavak A, Güngör H, et al.: Intermittent use of topical dimethyl sulfoxide in macular and papular amyloidosis. Int J Dermatol. 1998; 37(12): 949–54. PubMed Abstract \| Publisher Full Text\n\nVinnik CA, Jacob SW: Dimethylsulfoxide (DMSO) for human single-stage intraoperative tissue expansion and circulatory enhancement. Aesthetic Plast Surg. 1991; 15(4): 327–37. PubMed Abstract \| Publisher Full Text\n\nDonmez A, Tombuloglu M, Gungor A, et al.: Clinical side effects during peripheral blood progenitor cell infusion. Transfus Apher Sci. 2007; 36(1): 95–101. PubMed Abstract \| Publisher Full Text\n\nCordoba R, Arrieta R, Kerguelen A, et al.: The occurrence of adverse events during the infusion of autologous peripheral blood stem cells is related to the number of granulocytes in the leukapheresis product. Bone Marrow Transplant. 2007; 40(11): 1063–7. PubMed Abstract \| Publisher Full Text\n\nAlessandrino P, Bernasconi P, Caldera D, et al.: Adverse events occurring during bone marrow or peripheral blood progenitor cell infusion: analysis of 126 cases. Bone Marrow Transplant. 1999; 23(6): 533–7. PubMed Abstract \| Publisher Full Text\n\nBinnick SA, Shore SS, Corman A, et al.: Failure of dimethyl sulfoxide in the treatment of scleroderma. Arch Dermatol. 1977; 113(10): 1398–402. PubMed Abstract \| Publisher Full Text\n\nZuurmond WW, Langendijk PN, Bezemer PD, et al.: Treatment of acute reflex sympathetic dystrophy with DMSO 50% in a fatty cream. Acta Anaesthesiol Scand. 1996; 40(3): 364–7. PubMed Abstract \| Publisher Full Text\n\nGalmés A, Besalduch J, Bargay J, et al.: Cryopreservation of hematopoietic progenitor cells with 5-percent dimethyl sulfoxide at -80 degrees C without rate-controlled freezing. Transfusion. 1996; 36(9): 794–7. PubMed Abstract \| Publisher Full Text\n\nAmiridze N, Darwish R: Hemodynamic instability during treatment of intracranial dural arteriovenous fistula and carotid cavernous fistula with Onyx: preliminary results and anesthesia considerations. J Neurointerv Surg. 2009; 1(2): 146–50. PubMed Abstract \| Publisher Full Text\n\nVinokurov VL, Zharinov GM, Val’kovich AA, et al.: [The prevention of radiation injuries to the rectum and bladder in cervical cancer patients]. Vopr Onkol. 1990; 36(9): 1119–20. PubMed Abstract\n\nPercy EC, Carson JD: The use of DMSO in tennis elbow and rotator cuff tendonitis: a double-blind study. Med Sci Sports Exerc. 1981; 13(4): 215–9. PubMed Abstract\n\nWeigel BJ, Blaney SM, Reid JM, et al.: A phase I study of 17-allylaminogeldanamycin in relapsed/refractory pediatric patients with solid tumors: a Children’s Oncology Group study. Clin Cancer Res. 2007; 13(6): 1789–93. PubMed Abstract \| Publisher Full Text\n\nDawber R: Idoxuridine in herpes zoster: further evaluation of intermittent topical therapy. Br Med J. 1974; 2(5918): 526–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPerez RS, Zuurmond WW, Bezemer PD, et al.: The treatment of complex regional pain syndrome type I with free radical scavengers: a randomized controlled study. Pain. 2003; 102(3): 297–307. PubMed Abstract \| Publisher Full Text\n\nStav K, Beberashvili I, Lindner A, et al.: Predictors of response to intravesical dimethyl-sulfoxide cocktail in patients with interstitial cystitis. Urology. 2012; 80(1): 61–5. PubMed Abstract \| Publisher Full Text\n\nDuijvestein M, Vos AC, Roelofs H, et al.: Autologous bone marrow-derived mesenchymal stromal cell treatment for refractory luminal Crohn’s disease: results of a phase I study. Gut. 2010; 59(12): 1662–9. PubMed Abstract \| Publisher Full Text\n\nMarcacci G, Corazzelli G, Becchimanzi C, et al.: DMSO-associated encephalopathy during autologous peripheral stem cell infusion: a predisposing role of preconditioning exposure to CNS-penetrating agents? Bone Marrow Transplant. England, 2009; 44(2): 133–5. PubMed Abstract \| Publisher Full Text\n\nRabinov JD, Yoo AJ, Ogilvy CS, et al.: ONYX versus n-BCA for embolization of cranial dural arteriovenous fistulas. J Neurointerv Surg. 2013; 5(4): 306–10. PubMed Abstract \| Publisher Full Text\n\nMilone G, Mercurio S, Strano A, et al.: Adverse events after infusions of cryopreserved hematopoietic stem cells depend on non-mononuclear cells in the infused suspension and patient age. Cytotherapy. 2007; 9(4): 348–55. PubMed Abstract \| Publisher Full Text\n\nHoracek JM, Jebavy L, Jakl M, et al.: Cardiovascular changes associated with infusion of hematopoietic cell grafts in oncohematological patients -- impact of cryopreservation with dimethylsulfoxide. Exp Oncol. 2009; 31(2): 121–2. PubMed Abstract\n\nMurav’ev IuV: [Treatment of rheumatoid synovitis by intra-articular administration of dimethyl sulfoxide and corticosteroids]. Ter Arkh. 1986; 58(7): 104–5. PubMed Abstract\n\nAkkok CA, Holte MR, Tangen JM, et al.: Hematopoietic engraftment of dimethyl sulfoxide-depleted autologous peripheral blood progenitor cells. Transfusion. 2009; 49(2): 354–61. PubMed Abstract \| Publisher Full Text\n\nRössberger J, Fall M, Peeker R: Critical appraisal of dimethyl sulfoxide treatment for interstitial cystitis: discomfort, side-effects and treatment outcome. Scand J Urol Nephrol. 2005; 39(1): 73–7. PubMed Abstract \| Publisher Full Text\n\nBurton WJ, Gould PW, Hursthouse MW, et al.: A multicentre trial of Zostrum (5 percent idoxuridine in dimethyl sulphoxide) in herpes zoster. N Z Med J. 1981; 94(696): 384–6. PubMed Abstract\n\nLockie LM, Norcross BM: A clinical study on the effects of dimethyl sulfoxide in 103 patients with acute and chronic musculoskeletal injuries and inflammations. Ann N Y Acad Sci. 1967; 141(1): 599–602. PubMed Abstract \| Publisher Full Text\n\nEvstaf’ev VV: [The use of a heparin ointment in combination with dimexide in treating psoriasis]. Vestn Dermatol Venerol. 1989; (9): 71–2. PubMed Abstract\n\nParsons JL, Shepard WL, Fosdick WM: DMSO an adjutant to physical therapy in the chronic frozen shoulder. Ann N Y Acad Sci. 1967; 141(1): 569–71. PubMed Abstract \| Publisher Full Text\n\nDelaney C, Milano F, Cicconi L, et al.: Infusion of a non-HLA-matched ex-vivo expanded cord blood progenitor cell product after intensive acute myeloid leukaemia chemotherapy: a phase 1 trial. Lancet Haematol. 2016; 3(7): e330–9. PubMed Abstract \| Publisher Full Text\n\nStewart BH: Dimethyl sulfoxide (DMSO) in the treatment of troublesome genitourinary disorders: a preliminary report. Cleve Clin Q. 1966; 33(2): 81–4. PubMed Abstract\n\nOlver IN, Aisner J, Hament A, et al.: A prospective study of topical dimethyl sulfoxide for treating anthracycline extravasation. J Clin Oncol. 1988; 6(11): 1732–5. PubMed Abstract \| Publisher Full Text\n\nBrown JH: Clinical experience with DMSO in acute musculoskeletal conditions comparing a noncontrolled series with a controlled double blind study. Ann N Y Acad Sci. 1967; 141(1): 496–505. PubMed Abstract \| Publisher Full Text\n\nGarcia CA: Ocular toxicology of dimethyl sulfoxide and effects on retinitis pigmentosa. Ann N Y Acad Sci. 1983; 411: 48–51. PubMed Abstract \| Publisher Full Text\n\nKim DH, Jamal N, Saragosa R, et al.: Similar outcomes of cryopreserved allogeneic peripheral stem cell transplants (PBSCT) compared to fresh allografts. Biol Blood Marrow Transplant. 2007; 13(10): 1233–43. PubMed Abstract \| Publisher Full Text\n\nSalim AS: Role of oxygen-derived free radical scavengers in the treatment of recurrent pain produced by chronic pancreatitis. A new approach. Arch Surg. 1991; 126(9): 1109–14. PubMed Abstract \| Publisher Full Text\n\nMartino M, Morabito F, Messina G, et al.: Fractionated infusions of cryopreserved stem cells may prevent DMSO-induced major cardiac complications in graft recipients. Haematologica. 1996; 81(1): 59–61. PubMed Abstract\n\nEisenberg S, Wickline M, Linenberger M, et al.: Prevention of dimethylsulfoxide-related nausea and vomiting by prophylactic administration of ondansetron for patients receiving autologous cryopreserved peripheral blood stem cells. Oncol Nurs Forum. 2013; 40(3): 285–92. PubMed Abstract \| Publisher Full Text\n\nGonella S, Berchialla P, Bruno B, et al.: Are orange lollies effective in preventing nausea and vomiting related to dimethyl sulfoxide? A multicenter randomized trial. Support Care Cancer. 2014; 22(9): 2417–24. PubMed Abstract \| Publisher Full Text\n\nKessinger A, Schmit-Pokorny K, Smith D, et al.: Cryopreservation and infusion of autologous peripheral blood stem cells. Bone Marrow Transplant. 1990; 5 Suppl 1: 25–7. PubMed Abstract\n\nFoïs E, Desmartin M, Benhamida S, et al.: Recovery, viability and clinical toxicity of thawed and washed haematopoietic progenitor cells: analysis of 952 autologous peripheral blood stem cell transplantations. Bone Marrow Transplant. 2007; 40(9): 831–5. PubMed Abstract \| Publisher Full Text\n\nCalmels B, Lemarié C, Esterni B, et al.: Occurrence and severity of adverse events after autologous hematopoietic progenitor cell infusion are related to the amount of granulocytes in the apheresis product. Transfusion. 2007; 47(7): 1268–75. PubMed Abstract \| Publisher Full Text\n\nKeung YK, Lau S, Elkayam U, et al.: Cardiac arrhythmia after infusion of cryopreserved stem cells. Bone Marrow Transplant. 1994; 14(3): 363–7. PubMed Abstract" }	[ { "id": "40223", "date": "05 Feb 2019", "name": "Curly Morris", "expertise": [ "Reviewer Expertise haematology", "myeloma", "autologous transplantation" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nWith well over 20,000 patients receiving DMSO based autologous transplants annually in Europe alone, this is a timely review of the toxic effects of this valuable agent. It has been performed in an appropriate and scholarly manner and brings added value by including the Russian literature not easily accessible to the average English-speaking reader.\nHowever there are ways in which the review might be improved and give added value to the reader.\nIt is not easy to ascertain the number of patients receiving DMSO intravenously and those receiving it by other routes. A small table could clarify this.\n\nThe side effect tables either as numbers of patients or numbers of treatments. If they cannot be presented as one combined set of data then some explanation of the two separate tables would be beneficial.\nThere seems to have been no attempt to quantify the dose of DMSO which patients have received or to characterize the severity of the reactions and relate these. Furthermore DMSO is usually a vehicle to facilitate giving the patient some other treatment e.g. a stem cell transplant or drug so the reasons for the use of DMSO are not clear. This also means there are side effects from the drug or treatment facilitated by the DMSO; is it possible to separate these effects in any way? Do the authors of the many papers selected for analysis recommend an upper limit to the amount of DMSO given or have a strategy for minimising the dose?\nIn their final paragraph the authors suggest that \"reactions due to DMSO are often mild and transient\". In their previous paragraph they admit that the case reports the less common and more severe side effects which did not meet the eligibility criteria of this review. However as long ago as 2005 it was possible to identify severe side effects in an appreciable number of cases (Windrum et al., 20051). Furthermore although they do not separate the factors responsible the authors of reference 7 record a SAE (Grades 3, 4 and 5) profile in excess of 3%. The authors should possibly be a little more circumspect in this paragraph particularly as they recommend the use of DMSO in (unspecified) small doses.\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes\n\nIs the statistical analysis and its interpretation appropriate? Not applicable\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes", "responses": [ { "c_id": "4792", "date": "06 Aug 2019", "name": "bennedikte Madsen", "role": "Author Response", "response": "Dear Curly Morris, Thank for reviewing our manuscript: “Adverse reactions of dimethyl sulfoxide in humans: a systematic review”. We appreciate the effort put into reviewing our manuscript, and we have tried our best to use your comments to improve our manuscript. We have addressed the individual questions in the section below. We hope your find our replies satisfactory. Questions are written in italic and answers in plain. Q1: It is not easy to ascertain the number of patients receiving DMSO intravenously and those receiving it by other routes. A small table could clarify this. A1: We have added a table (Table 9) to our manuscript describing the route of administration of DMSO. Q2 & 3: There seems to have been no attempt to quantify the dose of DMSO which patients have received or to characterize the severity of the reactions and relate these. Furthermore DMSO is usually a vehicle to facilitate giving the patient some other treatment e.g. a stem cell transplant or drug so the reasons for the use of DMSO are not clear. This also means there are side effects from the drug or treatment facilitated by the DMSO; is it possible to separate these effects in any way? Do the authors of the many papers selected for analysis recommend an upper limit to the amount of DMSO given or have a strategy for minimising the dose? In their final paragraph the authors suggest that \"reactions due to DMSO are often mild and transient\". In their previous paragraph they admit that the case reports the less common and more severe side effects which did not meet the eligibility criteria of this review. However as long ago as 2005 it was possible to identify severe side effects in an appreciable number of cases (Windrum et al., 20051). Furthermore although they do not separate the factors responsible the authors of reference 7 record a SAE (Grades 3, 4 and 5) profile in excess of 3%. The authors should possibly be a little more circumspect in this paragraph particularly as they recommend the use of DMSO in (unspecified) small doses. A2 & 3: DMSO is most often used as a vehicle in combination with other drugs. Therefore, it is not possible to separate completely the adverse reactions related to the use of DMSO and the adverse reactions related to other drugs, since adverse reactions such as nausea, vomiting, headache etc. are not specific for solely DMSO. As described by the authors of reference 77, it was difficult to isolate the effect of DMSO from side effects related to conditioning chemotherapy. The only adverse effect that can solely be attributed to DMSO is halitosis. Therefore, we could not conclude that DMSO was the cause of SAE’s in reference 7 and have not included it in our study. Correctly, Windrum et al.1 describes several adverse reactions which may be contributed to DMSO. However, the study does not describe the seriousness of the adverse reactions. As described in our study, it is very possible that some events are underrepresented in our study, which is a limitation. The upper limit was not described by any studies; on the other hand several studies evaluated different doses of DMSO and found that a lesser amount of DMSO created fewer adverse reactions ( 61,86,120–122.). Based on this observation, we feel confident that the use of small amounts of DMSO is recommendable, since DMSO works well as a vehicle. However, limiting the amount would always be desirable." } ] }, { "id": "45643", "date": "21 Mar 2019", "name": "Igho Onakpoya", "expertise": [ "Reviewer Expertise Adverse drug reactions", "systematic reviews" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors have conducted a systematic review assessing reports of adverse reactions attributed to DMSO. The topic is interesting, and the authors have conducted their searches in a reasonable way. However, there are several flaws in this manuscript that need to be addressed:\nIntroduction\nThe term “possible adverse reactions” is incorrect. Suspected adverse reactions is more reasonable\nMethods\nIf Russian articles were screened by only one author, how were discrepancies resolved in these cases? Specify which authors extracted the data, and whether this was done independently.\nResults\nThe term “possibly due” is incorrect. There are 4 levels in describing associations between medicines and suspected adverse reactions. The authors should revise their terminology. You state “in some studies patients discontinued treatments due to halitosis”; however, you have provided references for 5 studies – the report can be more precise.\nDiscussion\nHow does “including Russian studies” strengthen the review? What about several other languages that have been omitted? You state that there seems to be a dose-response relationship, and have drawn similar conclusions. However, at no point in the results do you report data to support this claim. You state that studies reported associations between dose and the occurrence of adverse reactions, but fail to report the doses in question. Please enumerate the limitations of your review.\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Partly\n\nIs the statistical analysis and its interpretation appropriate? Partly\n\nAre the conclusions drawn adequately supported by the results presented in the review? Partly", "responses": [ { "c_id": "4793", "date": "06 Aug 2019", "name": "bennedikte Madsen", "role": "Author Response", "response": "Dear Igho J. Onakpoya, Thank your for reviewing our manuscript “Adverse reactions of dimethyl sulfoxide in humans: a systematic review”. Your comments were very helpful and we appreciate the effort you put in to reviewing our manuscript. We have addressed the individual questions in the section below. We hope your find our replies satisfactory. Questions are written in italic and answers in plain. Q1: The term “possible adverse reactions” is incorrect. Suspected adverse reactions is more reasonable. A1: We have changed the paragraph in the introduction section to “suspected adverse reactions”. Q2: If Russian articles were screened by only one author, how were discrepancies resolved in these cases? Specify which authors extracted the data, and whether this was done independently. A1: We have clarified in the manuscript how the screening process was performed: “Two authors (B.K.M. and D.Z.) independently screened title and abstract according to the eligibility criteria using www.covidence.org. Discrepancies were resolved by discussion. One author screened the full-text articles (B.K.M.). Russian articles were screened by an author fluent in Russian (M.H.). If M.H was in doubt regarding inclusion of a study the results were presented to B.K.M and then discussed until a mutual decision was made. After the screening process was finished, all included studies were imported to an Excel sheet (Microsoft Excel 2016). Data extraction was performed by two authors (M.H. extracted from the Russian articles and B.K.M. extracted from the rest).” Q3: The term “possibly due” is incorrect. There are 4 levels in describing associations between medicines and suspected adverse reactions. The authors should revise their terminology. A3: We have rewritten the paragraph so it now states: “Gastrointestinal adverse reactions were reported in 61 studies. Of these, 10 studies were randomized controlled trials.” Q4: You state “in some studies patients discontinued treatments due to halitosis”; however, you have provided references for 5 studies – the report can be more precise A4: We have made our report more precise and it now states: “In five studies, patients discontinued treatment due to halitosis.” Q5: How does “including Russian studies” strengthen the review? What about several other languages that have been omitted? A5: A Russian Chemist, Dr. Alexander Saytzeff, identified DMSO in 1866, however it was not used for medical use at the time1 . But the fact that he was Russian might have been the reason why Russian scientists have made numerous studies using DMSO. We therefore thought it would be valuable to include Russian articles since many of these studies have never been translated, and therefore are not available to the international scientific society. Of course, we could have included many other languages, but we thought it most relevant to include Russian since we observed a large amount of articles in Russian during our initial examination of the subject. Q6: You state that there seems to be a dose-response relationship and have drawn similar conclusions. However, at no point in the results do you report data to support this claim. You state that studies reported associations between dose and the occurrence of adverse reactions but fail to report the doses in question. A6: As mentioned in our study several studies described a dose-response relationship between the amount of DMSO and the occurrence of adverse reactions (26,33,41,53,67,71,73,75,78,83,85,86,93,101,115 ). However, since the doses of DMSO and the route of administration differ between the studies, we were not able to give an exact dose. We can only say that an association seems likely. Q7: Please enumerate the limitations of your review A7: We have enumerated the limitations listed in the discussion in the manuscript." } ] } ]	1	https://f1000research.com/articles/7-1746
https://f1000research.com/articles/8-86/v1	22 Jan 19	{ "type": "Research Article", "title": "Perceptions and use of technology in older people with ophthalmic conditions", "authors": [ "Zaria C. Ali", "Savana Shakir", "Tariq Mehmood Aslam", "Zaria C. Ali", "Savana Shakir" ], "abstract": "Background: Technologies such as mobile applications are increasingly being developed for patients to help manage their clinical conditions. However there is a paucity of information confirming the capacity or willingness of older patients with ophthalmic complaints to engage with such computer applications. The aim of this paper is to assess the perception and use of a range of common computing technologies by older ophthalmic patients, in order to guide future ophthalmology-specific development and clinical use. Methods: Patients attending Manchester Royal Eye Hospital were surveyed with questions designed to measure their perceptions, attitudes and experiences of using technology. Inclusion criteria included any patient aged 40 or over who attended the ophthalmology outpatients department. Results: A total of 300 patients completed the questionnaire. The male-to-female ratio was 169:127. The majority of patients owned predominantly mobile forms of technology such as tablets and smart phones. The most common uses of technology were for communicating with friends, watching television and gathering information. Patients aged over 80 had particular difficulty using technology and used it less regularly. Less than 10% overall stated eyesight as a reason for stopping using technology. Conclusions: Technology is used regularly by a large proportion of older ophthalmic patients, with numbers reducing significantly only in those aged 80 years or over. There appears to be potential for further medical use, though developers and clinicians should consider the perceptions and challenges highlighted through this survey.", "keywords": [ "Ophthalmology", "Perceptions", "Opinions", "Technology", "Elderly" ], "content": "Introduction\n\nOver the past decade there has been a surge of in the number of digital technologies aimed at assisting patients and health professionals in managing clinical conditions, including applications (apps) to help monitor chronic diseases1 and social media to share experiences and information2. Much of this new technology focuses on diabetes, hypertension and weight management1,3–8, but there have been developments in a wide range of clinical conditions3,9, including ophthalmology.\n\nOne large area of interest is in apps on computer tablets or phones to monitor vision at home to benefit those with conditions such as age-related macular degeneration (AMD)10,11. Other applications include those to help educate patients on their nutrient intake or ocular condition12–14. Ophthalmic patients may have distinct characteristics relative to other clinical groups in terms of their epidemiology, clinical conditions and treatment burdens. Whilst there are many devices available and in development for ophthalmic patients, there is relatively little information on the specific needs, preferences and perceptions ophthalmic patients may have towards these emerging technologies. Although there are studies looking at how patients in general use and feel about technology15,16, there is a distinct lack of studies looking at how patients specifically with ophthalmic conditions utilise it. The importance of understanding the intended patient bases with any technology development is highlighted by those trying to integrate web based tools for those with mental health problems. Some studies have found that negative views of users lead to disuse17, with target populations unwilling to engage18.\n\nThe aim of this study was to assess how patients with ophthalmic conditions use technology and their attitude towards it, with a focus on older patients recruited from outpatient clinics.\n\n\nMethods\n\nThe study took place between November 2014 and July 2015. A survey was designed that would capture basic patient demographics, type of technology used, frequency of use, how patients used technology, their views regarding how useful technology is to them and what potential barriers are to using technology. Input for the questions were derived from a patient public involvement (PPI) group where participants were asked about their views regarding technology and how they felt about using technology to help them manage their ophthalmic health. In addition input was derived from consultant ophthalmologists and experts in clinical technology use. Ethical approval for the study was given by the ethics committee of London Camden and Islington (REC reference number14/LO/1496, IRAS ID159394), and research was conducted in accordance with the Declaration of Helsinki. The ethics committee felt written consent was not needed for this study, as such verbal informed consent was taken from all participants and taken as confirmed if the participant returned their questionnaire.\n\nThe resulting survey was issued to 300 patients who attended Manchester Royal Eye Hospital outpatient clinics. Although our main focus was to determine characteristics of patients over 50 years, we decided to include some younger patients as a comparison. Inclusion criteria was any person who was a patient at Manchester Royal Eye Hospital aged over 40 years. The only exclusion criteria was the inability to understand written English. Patients were recruited on an opportunistic basis; they were approached whilst waiting for their scheduled appointment. In order to gain the views of those with a variety of ophthalmic conditions patients from a variety of subspecialty clinics were approached, including oculoplastics, neuro-ophthalmology, medical retina, and vitreo-retinal and glaucoma clinics. Patients were offered assistance in completing the questionnaire or could complete it independently. A researcher was also readily available should the patient require further clarification for any of the questions.\n\nThe questionnaire and the rationale for each question is discussed below.\n\nQ1. Do you currently use technology to help with tasks in your everyday life, e.g. Mobile phones, computers, etc.?\n\nPatients could answer yes or no to this question. This was asked to give a quick and immediate insight into how many patients used technology on an everyday basis.\n\nQ2. Which of the following devices do you own?\n\n- Desktop computer\n\n- Laptop computer\n\n- Tablet (iPad, Nexus, Windows surface etc.)\n\n- Smart phone (iPhone, HTC etc)\n\n- iPod or MP3 player\n\n- eBook Reader\n\n- Other (Please state)\n\nFor each device patient could choose one of four options; ‘own and use’, ‘own but don’t use’, ‘plan to buy’ and ‘don’t need’. This question was designed to assess which technology patients already owned and therefore which would be the most useful to develop aids for. Discussing technology within the PPI group revealed that although patients may own computers or tablets they may not use it which may cause the results to be misleading, so the option of ‘own but don’t use’ was included. Conversely, patients may be planning to purchase these devices in the future, so to survey this potential interest the option of ‘plan to buy’ was also included. The option of ‘don’t need’ was also added to see if patients felt they were unnecessary for them.\n\nQ3. How often do you use technology to help with the following activities?\n\n- Gathering Information e.g. researching a purchase or finding a recipe\n\n- Communicating with friends e.g. mobile phone, email, social media, online groups, Skype, texting etc\n\n- Listening to the radio or Watching TV shows or Videos\n\n- Finding information about your medical diagnosis, doctor, or healthcare organisation\n\n- Booking appointments e.g. Doctors, Opticians, Dentist\n\n- Creative pursuits – photography, music, genealogy etc.\n\n- Playing Games\n\n- Using online tests to test your health\n\n- Reading\n\n- Online learning e.g. classes to learn a language or new skill\n\n- Planning a travel route (e.g. via Satnav or public transport)\n\nFor each activity patients could choose one of 5 options:\n\n- Regularly (daily)\n\n- Often (a few times per week)\n\n- Sometimes (monthly or less)\n\n- I know of them\n\n- Never\n\nQuestion 4 was designed to gauge how widely used technology was in our patient base, and if it was purely for leisure, practical tasks such as planning a travel route, or health related tasks such as booking appointments or looking up information. The activities outlined above are the common tasks patients mentioned they carried out using technology during the PPI meeting. If patients used technology for practical and/or health related tasks already it may be they are more open to using technology to help manage their health. Giving the option of choosing how often they used technology for the various tasks also allowed us to see how regularly they used it.\n\nQ4. How much difficulty do you have using technology?\n\nPatients could choose one of four options:\n\n- No difficulty at all\n\n- A little difficulty\n\n- Moderate difficulty\n\n- Extreme difficulty\n\nA barrier to using technology could be difficulty in actually using it which was why this question was posed.\n\nQ5. How much experience do you have with the following?\n\n- Using Facebook\n\n- Using phone and/or tablet apps\n\n- Reading eBooks\n\n- Using Twitter\n\n- Sending email\n\n- Browsing the internet\n\n- Watching online TV or videos (iPlayer, YouTube etc.)\n\n- Listening to podcasts or online radio broadcasts\n\n- Using Skype\n\n- Playing games\n\n- Other (please state)\n\nFor each activity patients could choose one of four options; they could either state they were an ‘expert’, ‘amateur’, ‘novice’ or that they ‘never use’ technology for this purpose. It has previously been found that although patients may use technology on a regular basis they may perceive their ability to use technology as poor15. We therefore thought it valuable to assess how they felt about using some of the most popular technology currently used, as this could affect their willingness to engage in aids utilising these platforms.\n\nQ6. Does anything stop you from using some of the technology we have mentioned?\n\n- I don't have time to learn how to use it\n\n- My ICT skills are poor\n\n- Devices are too complicated to learn\n\n- My eyesight is too bad to see clearly\n\n- Technology is not for people like me\n\n- I am not aware of technology and what it can be used for\n\n- It is too expensive\n\n- It is too invasive - I don't want to use it\n\n- It is physically too hard to use\n\n- Other reason\n\nPatients could choose one of the following options for each statement:\n\n- Strongly agree\n\n- Mostly agree\n\n- Neither\n\n- Mostly disagree\n\n- Strongly disagree\n\nThis final question was included to try and ascertain what could be barriers to patient’s use of technology. This questions includes answers addressing patient’s attitudes towards technology that may be more amenable to change with good communication and user friendly and intuitive programs. It also included practical issues such as expense that may be able to be addressed with funding or by focusing on technology that patients tend to already own rather than developing completely new systems. Finally it includes physical barriers such as poor eyesight, or technology being physically too hard to use e.g. for those with severe arthritis. This may be more difficult to overcome, but could highlight the need to have other options such as enlarged text available.\n\n\nResults\n\n300 patients were recruited. Male-to-female ratio was 169:127 (3 did not specify their gender). The greatest frequency of patients (44%) were in the group aged 66–79 years. Participants’ ages are summarised in Table 1.\n\nAll were patients attending Ophthalmology outpatient clinics at the time of recruitment and 76.7% reported that they felt they had active ophthalmic problems. 32% had AMD, 14.7% had glaucoma and 4.7% were referred with cataracts. 26% did not specify what ophthalmic condition they had and answered ‘other’. 17.3% reported having no diagnosis. The raw, underlying data is available on OSF19.\n\nQ1. Do you currently use Technology to help with tasks in your everyday life, e.g. Mobile phones, computers etc.?\n\nOverall 66% of patients stated they own and used technology to help with everyday tasks. Results for each age group is shown in Table 2.\n\nQ2. Which of the following devices do you own?\n\nResults are shown in Figure 1. The four most commonly owned devices were smart phones, laptops and tablets.\n\nQ3. How often do you use technology to help with the following activities?\n\nTable 3 summarises the percentage of patients who use technology for the various activities. ‘Used’ was defined as those who answered that they either did that activity regularly, often or sometimes. The age groups were able to be grouped as it was found the most commonly conducted activities were the same in all age groups, with the three most common activities being communicating with friends, watching TV, and gathering information. The results for each individual age group and for each separate response can be found in the Extended data, Supplementary material 120. The main differences between age groups was a greater proportion of those under 65 stated they used technology to communicate with friends (77%) compared to those over 65 (36%). Those under the age of 65 were more likely to research their medical condition (17.5%) compared to those over the age of 65 (7.5%). Less than 10% of each age groups used technology to book appointments, do creative pursuits, do health tests or play games.\n\nQ4. How much experience do you have with various types of technology?\n\nOverall, participants were most comfortable sending emails and browsing the internet. The results for all age groups highlighting the two responses ‘expert’ and ‘never use’ are summarised in Table 4.\n\nThe results for all responses for each age group for each response can be found in the Extended data, Supplementary material 221. In age groups below the age of 80, at least a third of all patients stated they were experts at these two activities. In those aged over 80, less than 10% cited feeling they were an expert in the use of any type of technology.\n\nThe majority stated they did not use social media such as Facebook (70% overall did not use it) and twitter (90% overall stated they did not use it).\n\nQ5. How much difficulty do you have using technology?\n\nThe degree of difficulty experienced by patients is shown in Figure 2.\n\nFrom the age of 50 onwards, a majority of patients felt they had difficulties with using technology. This clearly increased with increasing age, with 13.9% of those over the age of 80 having extreme difficulties.\n\nQ6. Does anything stop you from using some of the technology we have mentioned?\n\nResults of more specific barriers to using technology are summarised in Figure 3.\n\n\nDiscussion\n\nThis survey provides detailed information on both the levels of technology use by older ophthalmic patients and the challenges they face, highlighting differences between different age groups.\n\nUse of technology is ubiquitous amongst age groups of 40–49 years in their everyday lives. It is interesting to note this remains at a near-universal level till the ages of 66 years and above, and even then it drops a small amount to 88.5%. A more significant drop to 58% is seen only once the age group of over 80s is reached. The association between likelihood of using technology and age has been demonstrated previously in a survey of 1371 cancer patients where a significant correlation between increase in age and likelihood of using technology was found16.\n\nAlthough the uptake and use of technologies for older age groups, especially the over 80s tends to reduce, there is no indication that this is perceived as due to inherent incapability or levels of expressed difficulty. Causes that were advocated as challenges were those that could potentially be addressed by education, with poor IT skills and over-complex devices cited as key barriers. Poor vision, expense and invasiveness were less important.\n\nMost of the activities that patients engage in using technology are not connected to their medical condition, including watching TV and communication with friends, especially with the relatively younger ages (under 65). Although a large proportion of patients utilised technology to gather information, less than 10% overall cited using technology to research their condition. It has been found that issues with using technology to research medical conditions could be connected to how trustworthy patients perceive the source to be8. Technology which has been advocated and developed by health professionals and has verified information could be highly valued22. Universally patients rarely used technology to book appointments or to do health tests which may be due to a lack of health services offering this service. The ability to do this could be valued, if not by the patient but by their carers, as demonstrated by a qualitative study whereby carers felt they benefited from the use and access of a health care portal23. There is clearly further scope in older patients for using the widespread ownership and ability to use portable devices for medical purposes such as patient education, interaction or monitoring. This may need at least initially to involve use of emails and websites. Developments which may not be as useful may be those revolving around social media. Although many reported regularly using technology to communicate with friends, patients generally did not feel as confident using common platforms such as Facebook and Twitter. Although social media has previously been found to help share information and experiences in those with HIV2, this sort of tool may not be the most appropriate for this patient group as they may not engage. The disinterest in communicating with others was also found by Girault et al., where only 54% cited communicating with peers as important16. This was an interesting contrast to other studies, where a feeling of connecting with others was one of the more positive elements of the technology related intervention2,4.\n\nThe most commonly owned technologies are portable devices; smart phones, laptops and tablet computers. Development of technologies on these devices would allow them to be accessible to a majority of patients up to the age of 66. They therefore could easily access programmes such as apps. which in turn could help improve adherence to treatment regimens4 or help monitor micro-nutrient intake to help prevent the development and progression of conditions such as AMD24,25. However, beyond this age, computer device ownership drops appears to drop. Programmes to help ophthalmic patients with apps on such devices might need to incorporate distribution of such devices to older generations. In ophthalmic patients, written text is more likely to be accessible if designed for tablet computers and smart phones than e-book readers, which, despite their potential for use in the ophthalmic community, are still less popular than smart phones or tablets.\n\nIn general the uptake and use of technological devices by older patients is high. It drops in patients over the age of 80, but it is likely that this can at least be combatted by better education and training and targeting software to devices in widespread use.\n\nComfort and confidence using technology appears to be an important obstacle to take into account when developing new technology for older ophthalmic patients. The younger patient groups who cited using technology regularly were also the ones who stated minimal difficulty using it. Comparatively those aged 66 and over were more likely to report having difficulty using technology and felt their ICT skills were poor despite using it regularly. This was also found in a previously published survey of 255 older patients in the community aged 60 or over15. If patients perceive the technology to be difficult to use they may not use it particularly as over a third of all patient groups felt technology was too invasive and nearly a half of those over 80 feeling technology ‘wasn’t for them’.\n\nEnsuring any new technology developments are explained properly and ensuring patients have a source of support to help them understand and troubleshoot any potential problems could help overcome any trepidation in using technology. In all age groups bar the over 80s, at least a third cited being ‘experts’ at sending emails and browsing the internet, so technology support could be based online either via e-mails or a help forum. Furthermore the majority did not agree with the sentiment that finding time to use technology was a drawback which suggests that they may be willing to invest time to learn to use it. Indeed patients are more likely to use technology once they are familiar and comfortable with it26.\n\nAnother issue is expense, as a notable proportion of all age groups felt this was a barrier to technology use. Creating technology which focuses on what patients currently own could increase the likelihood of patients engaging with it. It is also important to be aware that technology cannot completely replace a health professional and patients have frequently cited the importance of in-person communication1,4,23,27.\n\nOur study has a large sample size, with the majority suffering from ophthalmic conditions so we can be confident that the findings are reflective of this patient group. However there are limitations to our study. Although most had ophthalmic conditions, nearly a quarter of patients felt they did not. This is likely due to the fact that these patients attended for follow up at a general clinic following an acute eye problem.\n\nWe also did not explore other demographic data which may explain how patients use technology. Our study did not gather data on ethnic origin or social class, which might have provided further useful information for application development. For example, a study exploring the use of a health portal system found that those with lower health literacy and ethnic minorities are less likely to use health portals23. If this is similar in our patient base it may suggest that additional support such as the availability of alternative languages may be required. It may also have been of interest to enquire as to whether they lived alone as one study looking at the acceptability of e-health interventions in chronic pain found those who were older and lived alone were more likely to use technology28. It may be that those who were older in our patient group already had family or relatives to support them nearby, so did not feel the need to use technology to help them.\n\nIn our survey participants often didn’t answer questions which may have skewed results; for example when answering question 4 they may not have answered sections as they did not use technology for that particular activity, meaning the answer should have actually been ‘never’. There may also have been some difficulties understanding the difference between the options of ‘amateur’ and ‘novice’ in question 6. If further studies were carried out it may be of benefit to have the researcher sit and complete the questionnaire with the participant to answer any queries. Indeed this was done with 55 patients and resulted in these questionnaires being completed fully. It appears younger patients answered questionnaires more fully, so more additional support may only be required for older patients.\n\nExploring the attitudes of health care providers towards using technology with patients could also be important. A comparative survey of 1406 health providers and 1102 ‘consumers’ found that the consumers were more supportive of new medical technology9. It would be worth finding out the perspective of ophthalmic health professionals in using technology with their patients in daily practice as they would be the ones facilitating their use. Any new technology which is developed should also involve relevant health professionals as a mismatch between perceived benefit and applicability may affect its use in the clinical setting29.\n\n\nConclusion\n\nOverall our patients were found to have and to use predominantly portable devices such as smart phones and tablets suggesting that new technology using these mediums could be easily accessed. Although the majority regularly use technology, many still feel under-confident with new technologies and may not perceive it as beneficial to them particularly in those aged over 65. It is therefore important that the benefit of any new technology is explained clearly to the intended patient base, is individualised, and patients carefully instructed in its use with access to support should they need it. Further studies looking at other potential barriers to using technology in detailed socioeconomic and cultural groups may be of use and it may also be of value to collect health professionals’ views towards using technology with their patients.\n\n\nData availability\n\nOriginal data is available via figshare under the title ‘Original data MANAGER1’. DOI: https://doi.org/10.6084/m9.figshare.7358987.v119.\n\nSupplementary material 1. Graphs depicting full results for question 3: ‘How often do you use technology to help with the following activities?’ Extended data for MANAGER1. DOI: https://doi.org/10.6084/m9.figshare.7358969.v120.\n\nSupplementary material 2. Graphs depicting full data for question 4: How much experience do you have with various types of technology? Extended data for MANAGER1, part 2. DOI: https://doi.org/10.6084/m9.figshare.7358984.v121.", "appendix": "Grant information\n\nThis study was supported by Thea Pharmaceuticals.\n\n\nAcknowledgements\n\nThe results of this study were presented as a poster presentation at EURETINA annual meeting, 17th–20th September 2018 in Nice, France. Many thanks to Sara Robinson for her help in the initial design of the study.\n\n\nReferences\n\nSimon AC, Gude WT, Holleman F, et al.: Diabetes patients' experiences with the implementation of insulin therapy and their perceptions of computer-assisted self-management systems for insulin therapy. J Med Internet Res. 2014; 16(10): e235. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTaggart T, Grewe ME, Conserve DF, et al.: Social Media and HIV: A Systematic Review of Uses of Social Media in HIV Communication. J Med Internet Res. 2015; 17(11): e248. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMorrison LG, Hargood C, Lin SX, et al.: Understanding usage of a hybrid website and smartphone app for weight management: a mixed-methods study. J Med Internet Res. 2014; 16(10): e201. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeon N, Surender R, Bobrow K, et al.: Improving treatment adherence for blood pressure lowering via mobile phone SMS-messages in South Africa: a qualitative evaluation of the SMS-text Adherence SuppoRt (StAR) trial. BMC Fam Pract. 2015; 16(1): 80. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWharton CM, Johnston CS, Cunningham BK, et al.: Dietary self-monitoring, but not dietary quality, improves with use of smartphone app technology in an 8-week weight loss trial. J Nutr Educ Behav. 2014; 46(5): 440–4. PubMed Abstract \| Publisher Full Text\n\nBlock G, Azar KM, Romanelli RJ, et al.: Diabetes Prevention and Weight Loss with a Fully Automated Behavioral Intervention by Email, Web, and Mobile Phone: A Randomized Controlled Trial Among Persons with Prediabetes. J Med Internet Res. 2015; 17(10): e240. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCarter MC, Burley VJ, Nykjaer C, et al.: Adherence to a smartphone application for weight loss compared to website and paper diary: pilot randomized controlled trial. J Med Internet Res. 2013; 15(4): e32. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGlynn L, Casey M, Walsh J, et al.: Patients' views and experiences of technology based self-management tools for the treatment of hypertension in the community: A qualitative study. BMC Fam Pract. 2015; 16(1): 119. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBoeldt DL, Wineinger NE, Waalen J, et al.: How Consumers and Physicians View New Medical Technology: Comparative Survey. J Med Internet Res. 2015; 17(9): e215. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAslam TM, Parry NR, Murray IJ, et al.: Development and testing of an automated computer tablet-based method for self-testing of high and low contrast near visual acuity in ophthalmic patients. Graefes Arch Clin Exp Ophthalmol. 2016; 254(5): 891–9. PubMed Abstract \| Publisher Full Text\n\nWang YZ, He YG, Mitzel G, et al.: Handheld shape discrimination hyperacuity test on a mobile device for remote monitoring of visual function in maculopathy. Invest Ophthalmol Vis Sci. 2013; 54(8): 5497–505. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAli ZC, Silvioli R, Rajai A, et al.: Feasibility of Use of a Mobile Application for Nutrition Assessment Pertinent to Age-Related Macular Degeneration (MANAGER2). Transl Vis Sci Technol. 2017; 6(1): 4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nUKvisionstrategy: Glaucoma in Perspective: New app to educate patients on glaucoma. 2015. [cited 2017 31/07/2017]. [About 2 screens]. Reference Source\n\nRNIB: There's an app for that. 2017. [cited 2017 31/07/2017]. [About 5 screens]. Reference Source\n\nScanlon L, O'Shea E, O'Caoimh R, et al.: Technology Use and Frequency and Self-Rated Skills: A Survey of Community-Dwelling Older Adults. J Am Geriatr Soc. 2015; 63(7): 1483–4. PubMed Abstract \| Publisher Full Text\n\nGirault A, Ferrua M, Lalloué B, et al.: Internet-based technologies to improve cancer care coordination: current use and attitudes among cancer patients. Eur J Cancer. 2015; 51(4): 551–7. PubMed Abstract \| Publisher Full Text\n\nMusiat P, Goldstone P, Tarrier N: Understanding the acceptability of e-mental health--attitudes and expectations towards computerised self-help treatments for mental health problems. BMC Psychiatry. 2014; 14(1): 109. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWhealin JM, Seibert-Hatalsky LA, Howell JW, et al.: E-mental health preferences of Veterans with and without probable posttraumatic stress disorder. J Rehabil Res Dev. 2015; 52(6): 725–38. PubMed Abstract \| Publisher Full Text\n\nMann D, Riddell L, Lim K, et al.: Mobile Phone App Aimed at Improving Iron Intake and Bioavailability in Premenopausal Women: A Qualitative Evaluation. JMIR Mhealth Uhealth. 2015; 3(3): e92. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTieu L, Sarkar U, Schillinger D, et al.: Barriers and Facilitators to Online Portal Use Among Patients and Caregivers in a Safety Net Health Care System: A Qualitative Study. J Med Internet Res. 2015; 17(12): e275. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMoeller SM, Parekh N, Tinker L, et al.: Associations between intermediate age-related macular degeneration and lutein and zeaxanthin in the Carotenoids in Age-related Eye Disease Study (CAREDS): ancillary study of the Women's Health Initiative. Arch Ophthalmol. 2006; 124(8): 1151–62. PubMed Abstract \| Publisher Full Text\n\nTan JS, Wang JJ, Flood V, et al.: Dietary antioxidants and the long-term incidence of age-related macular degeneration: the Blue Mountains Eye Study. Ophthalmology. 2008; 115(2): 334–41. PubMed Abstract \| Publisher Full Text\n\nOlson KE, O’Brien MA, Rogers WA, et al.: Diffusion of Technology: Frequency of Use for Younger and Older Adults. Ageing Int. 2011; 36(1): 123–45. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHale K, Capra S, Bauer J: A Framework to Assist Health Professionals in Recommending High-Quality Apps for Supporting Chronic Disease Self-Management: Illustrative Assessment of Type 2 Diabetes Apps. JMIR Mhealth Uhealth. 2015; 3(3): e87. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCurrie M, Philip LJ, Roberts A: Attitudes towards the use and acceptance of eHealth technologies: a case study of older adults living with chronic pain and implications for rural healthcare. BMC Health Serv Res. 2015; 15(1): 162. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPals RA, Hansen UM, Johansen CB, et al.: Making sense of a new technology in clinical practice: a qualitative study of patient and physician perspectives. BMC Health Serv Res. 2015; 15(1): 402. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAli Z, Aslam TM: Original data MANAGER1. figshare. Dataset. 2018.\n\nAli Z, Aslam TM: Extended data for MANAGER1.docx. figshare. Figure. 2018.\n\nAli Z, Aslam TM: Extended data for MANAGER 1 part 2. figshare. Paper. 2018." }	[ { "id": "45508", "date": "25 Mar 2019", "name": "Jill Waalen", "expertise": [ "Reviewer Expertise Clinical biostatistics and Epidemiology (including survey methodology) in areas including digital medicine", "clinical biostatistics", "preventive medicine", "genetics", "hematology", "osteoporosis", "aging" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe study describes the results of a survey of 300 patients age 40 years and older attending the Manchester Royal Eye Hospital regarding their use of technology and reasons for not using it. While the use of a simple and short apparently non-validated questionnaire along with a convenience sample limits the value of the results, it nonetheless provides data in an area that is currently underrepresented in the literature, i.e., how technology may be accepted and used in populations with special considerations, like limited eyesight.\n\nMost of the following specific suggestions for improvement are minor:\n\nMethods, paragraph 2: More details are needed to described the administration of the questionnaire (some of which are actually alluded to in the Discussion), including whether the questionnaire had to be completed on site (and before seeing a provider) and whether help was systematically available to the survey takers. Also, with such a variety of ages and conditions, it would have been helpful to have some measure of eyesight impairment. Was there any attempt to ascertain this? Results, paragraph 1: Gender numbers do not add up. 169 male + 127 female + 3 did not specify =299 (not 300). Results, paragraph 1: It is stated that 44% of patients were in the 66-79 years age category. It would be helpful to give the full breakdown by age, with n’s included, perhaps in Table 1 (which is entitled “Age of patients”, but in fact is \"Use of Technology by Age Category”). Results, graphs: Although the graphs are an efficient way to convey the results (and tends to be preferred by readers) compared with tables, given the comment in the Discussion that participants often didn’t answer questions, which may have skewed results), it would be helpful to have the denominators for the percentages reported in the graphs.\n\nMinor editorial comments:\nDiscussion, paragraph 3: “the over 80s” should be changed to “participants over age 80” or something less colloquial. Discussion, paragraph 9: would add “relatively” to the description of the sample size as large.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4736", "date": "05 Aug 2019", "name": "Zaria Ali", "role": "Author Response", "response": "Response to reviewerMany thanks for your review. Please find a summary of the changes made in response to your comments.1. Methods, paragraph 2: More details are needed to describe the administration of the questionnaire (some of which are actually alluded to in the Discussion), including whether the questionnaire had to be completed on site (and before seeing a provider) and whether help was systematically available to the survey takers. Also, with such a variety of ages and conditions, it would have been helpful to have some measure of eyesight impairment. Was there any attempt to ascertain this? The methods have been amended to state when the patients had to complete the questionnaire (before their appointment) and the availability of help (patients aware researcher nearby should they require assistance, but also had the option to sit and complete the questionnaire with a researcher should they prefer). Unfortunately we did not have an item on the questionnaire which would ascertain the level of visual impairment, although we agree this could have been helpful, and would be helpful to include in future studies.2. Results, paragraph 1: Gender numbers do not add up. 169 + 127 + 3 did not specify =299 (not 300).The data has been reviewed and the numbers have been amended (should be 128 not 127). 3. Results, paragraph 1: It is stated that 44% of patients were in the 66-79 years age category. It would be helpful to give the full breakdown by age, with n’s included, perhaps in Table 1 (which is entitled “Age of patients”, but in fact is \"Use of Technology by Age Category”).The tables have been rearranged to be clearer, and as suggested we have included the actual number of patients in table 1 as opposed to just percentages.Results, graphs: Although the graphs are an efficient way to convey the results (and tends to be preferred by readers) compared with tables, given the comment in the Discussion that participants often didn’t answer questions, which may have skewed results), it would be helpful to have the denominators for the percentages reported in the graphs.Thank you for your comment. We agree that it is useful to assess the percentage of patients for each age group that did not answer. For figure 1 we felt that adding the percentage of each age group for each device would complicate the graph and make it more confusing, as such it is written in text instead. For figure 2 it was possible to easily demonstrate what percentage of patients in each age group did not answer. For Figure 3 all patients answered this questions so the graph has been left unaltered Minor editorial comments: Discussion, paragraph 3: “the over 80s” should be changed to “participants over age 80” or something less colloquial. This has now been changed, and also been changed further down in the discussion as well. Discussion, paragraph 9: would add “relatively” to the description of the sample size as large. This has now been added." } ] }, { "id": "43423", "date": "21 May 2019", "name": "Michel Michaelides", "expertise": [ "Reviewer Expertise Retinal Diseases" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a very valuable addition to the literature which is timely and needed. It is well planned and executed and should be highly cited and encourage further development in this avenue of research. There is increasing recognition that technology has a central role to play in improving visual function and such studies are an integral part of this process.\nIs the study design appropriate and is the work technically sound?\nThe study involves a straightforward survey of patient responses to questions on use of technology. It is not a validated questionnaire , but does not claim to be so. It uses entirely appropriate survey methodology to acquire information on patient perceptions of technology use.\n\nAre sufficient details of methods and analysis provided to allow replication by others?\nThe methods of survey collection are described and the questions also provided. The responses to these questions are provided in enough detail to provide the reader with a snapshot of patient perceptions. It should be noted that this cannot necessarily be translated to all populations globally and the paper accepts this. The validity is limited to the population under question and is not necessarily transferable to other populations. It is however still very useful information as real world data a patient population in a typical teaching hospital in U.K\n\nThere is very little existing information on this topic and therefore this paper is very important in informing doctors and developers on patient attitudes to technology.\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-86
https://f1000research.com/articles/8-1355/v1	05 Aug 19	{ "type": "Research Article", "title": "Molecular detection of Epstein-Barr virus in breast cancer among Sudanese female population: a case-control study", "authors": [ "Eiman S. Ahmed", "Lubna S. Elnour", "Emmanuel E. Siddig", "Rowa Hassan", "Eiman S. Ahmed", "Lubna S. Elnour", "Emmanuel E. Siddig" ], "abstract": "Background: Breast cancer is the most common cancer in women worldwide and in Sudan. Breast cancer occurs due to a multifactorial process and infection with an oncogenic virus has been recently investigated as a possible risk factor for breast cancer. For nearly two decades, studies have incriminated Epstein-Barr virus (EBV) in the etiology of breast cancer. However, the results are unconvincing, and their interpretation has remained a matter of debate. The aim of this study was to detect EBV in breast cancer biopsies obtained from Sudanese female patients. Methods: A descriptive, hospital-based, case-control study, conducted at Faculty of Medical Laboratory Science, University of Khartoum, Khartoum, Sudan. Archival blocks were obtained from 115 patients with breast cancer and 115 controls during the period between November 2016 till March 2017. Results: Among 115 breast cancer tissue specimens, EBV DNA was identified in 42/115 (36.5 %) samples and was not identified in 73/115 (63.5 %) tissue samples. The highest frequency of EBV detection was among 41–60 year-olds (23/42, 54.7 %), followed by 21–40 year-olds (12/42, 28.5 %) and 61–80 year-olds (5/42, 11.9 %). In the control group, the majority were diagnosed with fibroadenoma (70.4%), followed by fibrocystic changes (10.4%) and lactating changes (0.9%). Conclusion: The data obtained in this study demonstrated that EBV was present in a high percentage of our study population; however, the exact role of EBV in Sudanese breast cancer needs to be studied more in depth.", "keywords": [ "Breast cancer", "Invasive Ductal carcinoma", "Epstein-Barr virus" ], "content": "Abbreviations\n\nEBNA-4, Epstein–Barr virus nuclear antigen 4; EBV, Epstein-Barr virus; LMP1, latent membrane protein 1; MEC, mammary epithelial cell; PCR, polymerase chain reaction\n\n\nIntroduction\n\nEpstein-Barr virus (EBV) was discovered in the early 1960s by direct electron microscopy and it infects up to 95% of the adult human population, primarily in their early life. It usually remains silent and doesn’t cause any symptoms as the immune system doesn’t treat the virus like an invader following the initial infection. However, infection can lead to severe manifestations in later life1,2. This virus has been found to be the cause of many types of cancers including Burkitt’s lymphoma, Hodgkin’s disease, non-Hodgkin’s lymphoma, nasopharyngeal carcinoma, as well as Leiomyosarcoma arising in immunocompromised individuals3,4. The main mechanism by which the virus can induce cell transformation from normal to malignant cells is not yet understood and is still under investigation, although some researchers have related this to EBV viral protein effects5–8.\n\nBreast cancer is one of the most frequent types of cancers and is the leading cause of death in females globally9. Several factors can promote the development of breast cancer, including infections, especially oncogenic viruses that have been proven to be possible risk factors for the development of the cancer. These viruses include mouse mammary tumour virus (MMTV, causes breast cancer in mice), EBV and human papilloma virus (HPV)10–12. EBV DNA was identified previously in several studies using different molecular based techniques, including sequencing, polymerase chain reaction (PCR), in situ hybridization, immunohistochemistry and real-time PCR13.\n\nIn Sudan, breast cancer is a leading type of cancer in adults, along with leukemia, prostatic carcinoma, lymphoma and colorectal carcinoma14. Therefore, we aim in this study to detect EBV in breast cancer biopsies obtained from Sudanese female patients.\n\n\nMethods\n\nThis is a descriptive, hospital-based, case-control study, conducted at Faculty of Medical Laboratory Science, University of Khartoum, Khartoum City, Sudan, during the period of November 2016 until March 2017.\n\nSamples were collected from Soba Teaching Hospital and Military Hospital in Khartoum state, Sudan, during the period of November 2016 to January 2017. Archival blocks of 115 female patients diagnosed with breast cancer and 115 normal breast lesions (controls) were retrieved for the detection of EBV. All archival blocks from female patients diagnosed with breast cancer were included in the study, with no further inclusion or exclusion criteria. For the controls, archival blocks from women with a breast lesion, not diagnosed as breast cancer, taken during the same period as the breast cancer patient samples were included. The control samples were not matched to the breast cancer patient samples. Analysis of these blocks were carried out during the period of February 2017 to March 2017.\n\nFormalin-fixed paraffin-embedded blocks (FFPB) were cut at a thickness of 10 μm using a rotary microtome and the ribbons of the cut tissues transferred into 1.5 ml eppendorf tube using sterile forceps. The samples were deparaffinized by adding 1 ml of xylene to each sample and incubating for 30 minutes to allow the removal of the wax from the tissue sections. This step was repeated twice. The tissue was rehydrated using a series of ethanol washes, using 95%, 80% and then 70% ethanol for 15 min for each.\n\nThe DNA was extracted using the phenol-chloroform method with the aid of lysis buffer that composed from Sodium dodecyl sulfate (SDS) and proteinase K (40 µl) to get rid of proteins, as previously described15–17. To purify the DNA from the lysate, buffer saturated phenol was used. Then, in order to precipitate the DNA, 3M sodium acetate and isopropanol were added and this was spun in a centrifuge at 16000 rpm for seven mins to pellet the nucleic acid. This step was followed by washing in 70% ethanol to get rid from the salts and centrifugation at high speed to reconstruct the pellet again.\n\nDNA was resuspended in distilled water, quantified and stored at −20 °C. The final extracted DNA was checked for purity and quantity using a Nanodrop 1000 spectrophotometer. If the 260/280 nm absorption ratio was between 1.8-2, the DNA was considered to be pure. Purified DNA was used for PCRs.\n\nTo determine if the EBV DNA was present in our samples, DNA was amplified for detection of the EBV LMP-1 gene using forward 5- CCG AAG AGG TTG AAA ACA AA-3 and reverse 5- GTG GGG GTC GTC ATC ATC TC-3 primers. PCR was performed in a final reaction volume of 25μl, with 1 μl of template DNA, using iNtRON’s Maxime PCR PreMix Kit (i-Taq, Korea; Catalog No. 25025) according to manufacturer instructions. The following PCR conditions were used: 4 min at 95°C; then 35 cycles consisting of a denaturing step at 94°C for 1 min, an annealing step at 60°C for 1 min and a polymerization step at 72°C for 1 min; followed by a single incubation step for 10 min at 72°C. PCR was performed on an Aeris Thermocycler (AERIS-BG096, Esco Micro Pte.Ltd, Singapore).\n\nThe amplification product of 131 bp was separated on a 2% agarose gel stained with ethidium bromide and visualized under UV light using the UVCI-1100 gel imaging system (Major Science, UVCI-1100, Saratoga, CA, USA). A 100 bp DNA ladder was used and a band of 131 bp was considered as a positive (Figure 1).\n\nPCR products were loaded on 2% agarose gel containing ethidium bromide and visualized under UV light. Lanes 1 and 13 show the 100bp DNA ladder. Lanes 2 – 12 show different samples containing the PCR product of 131 bp.\n\nThe statistical analysis was performed using SPSS (version 16.0) statistical software. Descriptive statistics were conducted to generate graphical and numerical summaries. The Chi-squared test was used to compare the frequencies of the categorical variables. A value of p < 0.05 was considered statistically significant.\n\nThe study was approved by the faculty research board of the faculty of Medical Laboratory Sciences, University of Khartoum, Sudan. Written informed consent was obtained from patients for the use of their samples and publication of results.\n\n\nResults\n\nThe age of the breast cancer patients ranged between 23 and 80 years and the mean age was 50.49 ± 9.50 years, while the control group age ranged from 25 to 72 years and the mean age was 37.65± 5.89 years18. Regarding histological type, 96 (83.5%) of the patients diagnosed with breast cancer had invasive ductal carcinoma, 18 (15.7%) had invasive lobular carcinoma and one (0.9%) had mucinous carcinoma. Regarding the clinical staging (according to the TNM staging of the American Joint Committee on Cancer), 10 (8.7%) were stage I, 13 (11.3%) were stage IIA, 15 (13%) were stage IIB, 17 (14.8) were stage IIIA, 24 (20.9%) were stage IIIB and 36 (31.1) were stage IV (Table 1). In the control group, the majority were diagnosed with fibroadenoma (70.4%), followed by fibrocystic changes (10.4%) and lactating changes (0.9%).\n\nAmong the 115 breast cancers tissue specimens, EBV DNA was identified in 42/115 (36.5 %) samples and was not identified in 73/115 (63.5 %) tissue samples (Table 2). In the control group, EBV was detected in 101/115 (87.8%) and only 14/115 (12.2%) tested negative for the presence of the virus. The Chi-square test showed that EBV has a significant effect on breast cancer with a p-value of 0.001.\n\nThe highest frequency of infection was seen in the age group 41–60, with infection found in 23/42 (54.7 %), followed by age ranges 21–40 (12/42, 28.5 %) and 61–80 (5/42, 11.9 %).\n\nNotably, with regard to EBV infection and staging of breast cancer, by far the majority of infections were detected among stage IV (11/42, 26.1%), followed by stage IIIA (10/42, 23.8%), and stage IIIB (8/42, 19.0%), although these results were shown to be statistically insignificant (Table 3).\n\n(Chi square test; p value =0.001).\n\n(Chi square test; p value > 0.05).\n\n\nDiscussion\n\nBreast cancer has been proven to be a leading cause of death amongst women in the western world after lung cancer. Its incidence has been confirmed to be higher in females than males, and the number of cases increase with age. Early case detection has improved due to the promotion of screening procedures and diagnosis19,20. In Sudan, breast cancer still accounts for many cases of cancer among women. In 1959, breast cancer was found to account 22.9% of all cancer cases. The bulk of those patients were females within the menopausal age, with disease clinically progressing in a similar pattern as in European ladies21,22. In 2010, an updated study of breast cancer prevalence in Sudan by Intisar and colleagues found a prevalence of 25.1 per 100,00023.\n\nFor approximately twenty years, studies have been conducted that have aimed to identify the association between EBV and breast cancer. The first report of an association between EBV and breast cancer was from a study that carried out screening of blood and DNA of 91 patients with breast cancer by PCR in order to detect the EBV genome in 199524. Since that, other studies have confirmed the finding that EBV is associated with breast cancer25–33. However, these results have been controversial since negative results were obtained in some studies and explanations of the results have been under discussion for several years25,34–38.\n\nMoreover, in 2016 Hu and colleagues queried whether the EBV can be associated with breast cancer development. NOD/SCID mice were used and their results suggested that EBV can infect primary human mammary epithelial cells (MECs) through CD21, leading to phenotypic variations consistent with transformation. These immortalized MECs infected with EBV cooperatively (with activated Ras) increase tumor formation in vivo, recapitulating a multi-step tumorigenesis in an established animal model39.\n\nA study was conducted in 2014 by Yahia and colleagues to study the relationship between breast cancer and the presence of EBV genome. Interestingly, this was the first study in Sudan to assess the role of EBV infection in the etiology of breast cancer. They used PCR and in situ hybridization to screen 92 breast cancer samples, aiming to measure the presence of Epstein-Barr nuclear antigen 4 (EBNA-4) and LMP-1. Their final outcome was statistically significant (p = 0.0001), demonstrating the presence of EBV genome in 49 (53.3%) and 10 (11%) patients by LMP-1 and EBNA-4 PCR, respectively10. This result is in agreement with our results, which showed that EBV DNA was identified in 42/115 (36.5 %) samples and was not identified in 73/115 (63.5 %) tissue samples.\n\nNotably, the incidence of EBV among Iranian females with breast cancer has been found to be low compared to our findings. In 2016, Mohammad Izadeh and colleagues tested LMP-1 antigen expression in breast carcinoma using immunohistochemistry in 80 Iranian patients. LMP-1 expression was detected in 6 cases (7.5%) of breast carcinoma cases and these results were found to be statistically significant. LMP-1 expression was not detected in the normal breast tissue, taken from a site adjacent to the carcinoma, in all cases40. Furthermore, another study from Iran, conducted by Kadivar and colleagues, to evaluate the presence of EBV in benign and malignant breast lesions using PCR and immunohistochemistry is also in disagreement with our results, as all breast cancer samples were negative for EBV, as were the control group samples41.\n\nThe present study verified the presence of EBV in 36.5% of the total breast cancer samples, in agreement with the study conducted by Bonet and colleagues, in which 51 (51%) cases were found to be positive for EBV by PCR out of a total of 100 breast cancer cases27.\n\nHowever, this study has some limitations. The study did not use cohort techniques to assess whether EBV viral load had an effect on the prognosis of breast cancer and we could not use advanced techniques for visualization of the virus in the malignant cells due to time and budget constraints.\n\n\nConclusion\n\nIn conclusion, the data obtained in this study demonstrated that EBV was present in a high percentage of our study population; however, the exact role of EBV in Sudanese breast cancer needs to be studied more in depth.\n\n\nData availability\n\nHarvard Dataverse: Replication Data for: EBV and Breast cancer Data among Sudanese patients. https://doi.org/10.7910/DVN/94YOCV18\n\nEBV_and_Breast_cancer_Data.tab (diagnoses, PCR results and ages of patients and controls)\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nWe would like to express our gratitude to Miss Abeer Musa for her indefatigable help and support.\n\n\nReferences\n\nAndrei G, Trompet E, Snoeck R: Novel Therapeutics for Epstein-Barr Virus. Molecules. 2019; 24(5): pii: E997. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCohen JI: Epstein-Barr virus infection. N Engl J Med. 2000; 343(7): 481–492. PubMed Abstract \| Publisher Full Text\n\nEdreis A, Mohamed MA, Mohamed NS, et al.: Molecular Detection of Epstein - Barr virus in Nasopharyngeal Carcinoma among Sudanese population. Infect Agent Cancer. 2016; 11: 55. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNiedboitek G, Young LS: Epstein-Barr virus persistence and virus-associated tumours. Lancet. 1994; 343(8893): 333–335. PubMed Abstract \| Publisher Full Text\n\nBenchimol S, Minden M: Virus oncogens and tumor suppressor genes. The basic science of oncology. Edited by: Tanno IFHR. New York: McGraw-Hill, 1998; 77–90.\n\nThomas M, Massimi P, Banks L: HPV-18 E6 inhibits p53 DNA binding activity regardless of the oligomeric state of p53 or the exact p53 recognition sequence. Oncogene. 1996; 13(3): 471–480. PubMed Abstract\n\nEltahir HA, Elhassan AM, Ibrahim ME: Contribution of retinoblastoma LOH and the p53 Arg/Pro polymorphism to cervical cancer. Mol Med Rep. 2012; 6(3): 473–476. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nXu M, Zhang WL, Zhu Q, et al.: Genome-wide profiling of Epstein-Barr virus integration by targeted sequencing in Epstein-Barr virus associated malignancies. Theranostics. 2019; 9(4): 1115–1124. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJemal A, Bray F, Center MM, et al.: Global cancer statistics. CA Cancer J Clin. 2011; 61(2): 69–90. PubMed Abstract \| Publisher Full Text\n\nYahia ZA, Adam AA, Elgizouli M, et al.: Epstein Barr virus: a prime candidate of breast cancer aetiology in Sudanese patients. Infect Agent Cancer. 2014; 9(1): 9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHennighausen L: Mouse models for breast cancer. Breast Cancer Res. 2000; 2(1): 2–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLawson JS, Heng B: Viruses and breast cancer. Cancers (Basel). 2010; 2(2): 752–772. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGlenn WK, Heng B, Delprado W, et al.: Epstein-Barr virus, human papillomavirus and mouse mammary tumour virus as multiple viruses in breast cancer. PLoS One. 2012; 7(11): e48788. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSaeed ME, Cao J, Fadul B, et al.: A five-year survey of cancer prevalence in Sudan. Anticancer Res. 2016; 36(1): 279–86. PubMed Abstract\n\nHilz H, Wiegers U, Adamietz P: Stimulation of proteinase K action by denaturing agents: application to the isolation of nucleic acids and the degradation of ‘masked’ proteins. Eur J Biochem. 1975; 56(1): 103–8. PubMed Abstract \| Publisher Full Text\n\nJoseph Sambrook DR: Molecular cloning: a laboratory manual. Cold Spring Harbor: Cold Spring Harbor Laboratory Press; 2001. Reference Source\n\nPikor LA, Enfield KS, Cameron H, et al.: DNA extraction from paraffin embedded material for genetic and epigenetic analyses. J Vis Exp. 2011; 49: pii: 2763. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRowa H: Replication Data for: EBV and Breast cancer Data among Sudanese patients. 2019. Harvard Dataverse, V2, UNF:6:rsC2x2VWcC9IyZzuBmOkHg== [fileUNF]. https://www.doi.org/10.7910/DVN/94YOCV\n\nIbrahim T, Mercatali L, Amadori D: A new emergency in oncology: Bone metastases in breast cancer patients (Review). Oncol Lett. 2013; 6(2): 306–310. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYousef EM, Tahir MR, St-Pierre Y, et al.: MMP-9 expression varies according to molecular subtypes of breast cancer. BMC Cancer. 2014; 14: 609. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLynch JB, Omar A: Cancer in the Sudan. Sudan Med J. 1963; 2: 29–37.\n\nElamin A, Ibrahim ME, Abuidris D, et al.: Part I: cancer in Sudan—burden, distribution, and trends breast, gynecological, and prostate cancers. Cancer Med. 2015; 4(3): 447–456. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSaeed IE, Weng HY, Mohamed KH, et al.: Cancer incidence in Khartoum, Sudan: first results from the Cancer Registry, 2009–2010. Cancer med. 2014; 3(4): 1075–84. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLabrecque LG, Barnes DM, Fentiman IS, et al.: Epstein-Barr virus in epithelial cell tumors: a breast cancer study. Cancer Res. 1995; 55(1): 39–45. PubMed Abstract\n\nArbach H, Viglasky V, Lefeu F, et al.: Epstein-Barr virus (EBV) genome and expression in breast cancer tissue: effect of EBV infection of breast cancer cells on resistance to paclitaxel (Taxol). J Virol. 2006; 80(2): 845–853. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMazouni C, Fina F, Romain S, et al.: Epstein-Barr virus as a marker of biological aggressiveness in breast cancer. Br J Cancer. 2011; 104(2): 332–337. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBonnet M, Guinebretiere JM, Kremmer E, et al.: Detection of Epstein-Barr virus in invasive breast cancers. J Natl Cancer Inst. 1999; 91(16): 1376–1381. PubMed Abstract \| Publisher Full Text\n\nFina F, Romain S, Ouafik L, et al.: Frequency and genome load of Epstein-Barr virus in 509 breast cancers from different geographical areas. Br J Cancer. 2001; 84(6): 783–790. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcCall SA, Lichy JH, Bijwaard KE, et al.: Epstein-Barr virus detection in ductal carcinoma of the breast. J Natl Cancer Inst. 2001; 93(2): 148–150. PubMed Abstract \| Publisher Full Text\n\nGrinstein S, Preciado MV, Gattuso P, et al.: Demonstration of Epstein-Barr virus in carcinomas of various sites. Cancer Res. 2002; 62(17): 4876–4878. PubMed Abstract\n\nXue SA, Lampert IA, Haldane JS, et al.: Epstein-Barr virus gene expression in human breast cancer: protagonist or passenger? Br J Cancer. 2003; 89(1): 113–119. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThorne LB, Ryan JL, Elmore SH, et al.: Real-time PCR measures Epstein-Barr Virus DNA in archival breast adenocarcinomas. Diagn Mol Pathol. 2005; 14(1): 29–33. PubMed Abstract \| Publisher Full Text\n\nPerkins RS, Sahm K, Marando C, et al.: Analysis of Epstein-Barr virus reservoirs in paired blood and breast cancer primary biopsy specimens by real time PCR. Breast Cancer Res. 2006; 8(6): R70. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGlaser SL, Ambinder RF, DiGiuseppe JA, et al.: Absence of Epstein-Barr virus EBER-1 transcripts in an epidemiologically diverse group of breast cancers. Int J Cancer. 1998; 75(4): 555–558. PubMed Abstract \| Publisher Full Text\n\nDadmanesh F, Peterse JL, Sapino A, et al.: Lymphoepithelioma-like carcinoma of the breast: lack of evidence of Epstein-Barr virus infection. Histopathology. 2001; 38(1): 54–61. PubMed Abstract \| Publisher Full Text\n\nDeshpande CG, Badve S, Kidwai N, et al.: Lack of expression of the Epstein-Barr Virus (EBV) gene products, EBERs, EBNA1, LMP1, and LMP2A, in breast cancer cells. Lab Invest. 2002; 82(9): 1193–1199. PubMed Abstract \| Publisher Full Text\n\nMurray PG, Lissauer D, Junying J, et al.: Reactivity with A monoclonal antibody to Epstein-Barr virus (EBV) nuclear antigen 1 defines a subset of aggressive breast cancers in the absence of the EBV genome. Cancer Res. 2003; 63(9): 2338–2343. PubMed Abstract\n\nHerrmann K, Niedobitek G: Lack of evidence for an association of Epstein-Barr virus infection with breast carcinoma. Breast Cancer Res. 2003; 5(1): R13–R17. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHu H, Luo ML, Desmedt C, et al.: Epstein-Barr Virus Infection of Mammary Epithelial Cells Promotes Malignant Transformation. EBioMedicine. 2016; 9: 148–160. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMohammadizadeh F, Zarean M, Abbasi M: Association of Epstein-Barr virus with invasive breast carcinoma and its impact on well-known clinicopathologic parameters in Iranian women. Adv Biomed Res. 2014; 3: 141. PubMed Abstract \| Free Full Text\n\nKadivar M, Monabati A, Joulaee A, et al.: Epstein-Barr virus and breast cancer: lack of evidence for an association in Iranian women. Pathol Oncol Res. 2011; 17(3): 489–92. PubMed Abstract \| Publisher Full Text" }	[ { "id": "52048", "date": "16 Dec 2019", "name": "Sulma Ibrahim Mohammed", "expertise": [ "Reviewer Expertise Cancer biology sp breast cancer." ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn the Abstract: Method: the age of the control group is far younger compared to patients (37 vs 50). Result section - control group - only the histological type is provided. It should show the virus presence in tissue of patients with breast cancer compared to control without cancer - that does not make sense.\nIn the Methods: how are the tissues fixed and embedded in blocks and how long?\nResults: Table 1 - describe only patients - why not control as well?\nTable 2: does not make sense and is not clear.\nIn results sections: last paragraph state that \"In the control group, EBV was detected in 101/115 (87.8%) and only 14/115 (12.2%) tested negative for the presence of the virus\" that means the virus is present in the majority of control samples - which render the conclusion wrong.\nFig 1 is not necessary unless control samples are included as well.\nDiscussion: the author should discuss their results and not cite the literature (this was done in the introduction) - they should discuss what their results mean and what was the importance of looking at the stage, age and virus presence.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [] }, { "id": "59388", "date": "04 Feb 2020", "name": "James Lawson", "expertise": [], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nUnfortunately the manuscript and the study have serious flaws.\n\nThe most important flaw is the lack of recognition of infiltrating lymphocytes in breast cancer specimens. These lymphocytes can be Epstein Barr positive and give false positive outcomes of PCR based studies.\n\nThis issue has long been recognised by others and is repeatedly referred to in the literature\n\nThis problem can be overcome in several ways as published.\n\nIn addition the results of the controls are not reported properly? See Table 2. This table does not make sense. This problem has been commented by a previous reviewer.\n\nI am sorry but this project is not of sufficient quality to deserve indexing.\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [] } ]	1	https://f1000research.com/articles/8-1355
https://f1000research.com/articles/8-891/v1	20 Jun 19	{ "type": "Method Article", "title": "A refinement to the formalin test in mice", "authors": [ "Douglas M Lopes", "Heather L Cater", "Matthew Thakur", "S.E. Wells", "Stephen B McMahon", "Heather L Cater", "Matthew Thakur", "S.E. Wells", "Stephen B McMahon" ], "abstract": "The constant refinement of tests used in animal research is crucial for the scientific community. This is particularly true for the field of pain research, where ethical standards are notably sensitive. The formalin test is widely used in pain research and some of its mechanisms resemble those underlying clinical pain in humans. Immediately upon injection, formalin triggers two waves (an early and a late phase) of strong, nociceptive behaviour, characterised by licking, biting, lifting and shaking the injected paw of the animal. Although well characterised at the behaviour level, since its proposal over four decades ago, there has not been any significant refinement to the formalin test, especially those combining minimisation of animal distress and preservation of behavioural outcomes of the test. Here, we propose a modified and improved method for the formalin test. We show that anaesthetising the animal with the inhalable anaesthetic sevoflurane at the time of the injection can produce reliable, robust and reproducible results whilst animal distress during the initial phase is reduced. Importantly, our results were validated by pharmacological suppression of the behaviour during the late phase of the test with gabapentin, the anaesthetic showing no interference with the drug. In addition, we demonstrate that this is also a useful method to screen for changes in pain behaviour in response to formalin in transgenic lines.", "keywords": [ "Pain", "Formalin", "Nociception", "Behaviour", "Refinement", "3Rs" ], "content": "Introduction\n\nSince first reported over 40 years ago, the formalin test1 has been widely used in pain research and is known to capture mechanisms that are likely to be relevant to many pain patients in clinic2,3, including the poorly localized, burning and throbbing pain sensation4. The unique feature of the formalin test is that it triggers two phases of nociceptive behaviour: the first one is directly linked to the stimulation of primary sensory neurons and is followed by a second phase, which is associated with inflammation and involves central sensitisation1,5–7. These phases are marked by striking pain behaviour in which the animals lick, bite, lift and shake the injected paw. Over many years, different groups have extensively characterised the test, demonstrating robust and reproducible quantitative behaviour outcomes5,8–12.\n\nThe constant refinement of experimental procedures involving animals is important for the whole research community but is particularly important for the field of pain research, due to the obvious ethical implications of this type of research. Regulations regarding the use of animals for research in the UK dates back to the 19th century, with strict safeguards to avoid or minimise animal suffering and cruelty, as well as ensuring high animal welfare standards are met13,14. Furthermore, the 3Rs concept (reduce, refine and replace) ensures that for every experiment, the use of animals is absolutely necessary13,15–17. Further to these ethical considerations, and of particular relevance to the formalin test, other evidence suggests that restraining also induces stress, behavioural and other physiological changes in the animals18–25, including hyperalgesia26–28, which can impact the outcome of the test. As highlighted in the above guidelines and given the current standard procedure for the formalin test, where the animals are physically restrained and may experience unnecessary levels of stress, the present study aimed to refine the current method used for the injection of formalin, without compromising its reproducibility. We posed the question of whether anaesthetising the animal at the time of formalin injection could result in more consistent injections and reduce the stress experienced by the animals, without losing the behavioural effects that formalin triggers.\n\nWe show that the use of an inhalable anaesthetic during the time of formalin administration minimises animal stress and improves injection consistency. Whilst the anaesthetic reduced the behaviours observed during the first response phase, it appears not to affect the responses observed during the second phase of the test. We validated our proposed method by showing its sensitivity to a known analgesic agent, gabapentin, as well as its efficacy in different transgenic mouse lines. Together, our data present a refined method for the formalin test, whilst also demonstrating that the second phase can occur without a behavioural response during the first phase.\n\n\nMethods\n\nAll experiments were performed in accordance with the UK Animals (Scientific Procedures) Act 1986 and Local Ethical Committee approval. All efforts were made to minimise the suffering of animals during the experiments by carefully following the procedures.\n\nIn all experiments, adult (11 to 13 weeks of age) homozygous and wildtype, male and female littermates were used. For the gabapentin and sevoflurane experiments, wildtype C57BL/6 mice were used. Studies on the anaesthetised groups were performed on animals bred at Mary Lyon Centre (MLC; Harwell, UK), whereas the remaining animals were bred at King’s College London (KCL; UK). For the experiments using transgenic animals, mice carrying the null alleles Pink1tm1b(EUCOMM)Wtsi and Slit1tm1b(Komp)Wtsi were generated at Harwell (UK), as part of the International Mouse Phenotyping Consortium (IMPC) and maintained as heterozygotes on a C57BL/6NTac background. The colony was intercrossed and genotyped by an independent experimenter, to ensure effective blinding during behavioural testing.\n\nAnimals were housed in IVC cages (Tecniplast – 1284L and 1285L, with autoclaved Datesand Aspen bedding) in groups of 2 to 5 per cage, under 12-hour-on/12-hour-off cyclic lighting (30minutes dusk to dawn, dawn to dusk period), at controlled temperature (21 ± 2°C) and humidity (55 ± 10%) conditions. Cage bases were changed weekly. The mice had free access to filtered water (25 p.p.m. chloride) and were fed ad libitum on a commercial diet - either Rat and Mouse Diet No. 3 [RM3] (Special Diet Services, Essex, UK), composed of 5.3% fat [corn oil], 21.2% protein, 57.4% carbohydrate and 4.6% fibre (provided at the MLC; Harwell BSU facility), or PicoLab Rodent Diet 5053 (LabDiet St. Louis, MO, USA), composed of 21% protein, 4.5% fat and 6% fibre (provided at King’s College London BSU facility). All mice housed in the same facility received the same food. Food was irradiated to 2.5 Mrads. All mice went through daily health checks for general physical and health appearance (e.g. coat, eyes and ears appearances, fighting wounds), bedding/water bottle appearance and any signs of distress. The MLC and KCL are specific pathogen free centres.\n\nA total of 95 mice were used for the experiments: Sevoflurane experiments: N = 16 anaesthetised and N = 8 non-anaesthetised; Gabapentin experiments: N = 20 gabapentin and N = 16 controls; Pink1 experiments: N = 8 Pink1 -/- and N = 5 controls and Slit1 experiments :N = 11 Slit1 -/- and N = 11 controls. Animals were allocated using simple random randomization, i.e. subjects to each group purely randomly for every assignment.\n\nMice were anaesthetised by inhalation of sevoflurane (5% flow) (Zoetis, UK) for 2 minutes, followed by subcutaneous injection of formalin (20ul at 1.85% concentration) (Sigma, UK, Cat Nº 252549) into the right hind paw. A 30-gauge needle was used to perform the injections. A single animal was then placed into the arena (details below, and in Figure 1) and camera recording was started. For the gabapentin experiments (experiments related to Figure 3), gabapentin (Sigma, UK, Cat. № G154) was injected intraperitoneally at a concentration of 50mg/kg, immediately prior to the injection of formalin. All experiments were performed during the light cycle (usual starting time 9am).\n\n(a) Perspex acrylic glass arenas (36 x 40 x 13 cm), consisting of three mirrored and one transparent wall, were used for the formalin test. The recording camera was placed facing the arena, at an approximate 30° angle to the arena surface and 50 cm away from the transparent wall, so all mouse behaviour could be easily recorded for later analysis. (b) Video still of an experiment, showing that the behaviour of the animal could be meticulously monitored and accessed.\n\n(a) Graph representing behavioural response to formalin injection over time. Anaesthetised animals which receive gabapentin (peach squares) present a diminished response to formalin injection in comparison to the control group (white diamonds; anaesthetic only). ANOVA repeated measures, time: F(3.8, 129.6): 7.998; p< 0.001; timegroup: F(3.8, 129.634): 6.92; p <0.001. (b) Bar graphs representing the area under the curve show no difference between the groups in the early phase of the test, but a clear effect of the drug during the late phase (p < 0.001; One-way ANOVA; N = 20 gabapentin and N = 16 control). Graphs represent means +/- SEM.\n\nPerspex acrylic glass arenas (36 x 40 x 13 cm), consisting of three mirrored and one transparent wall, were built in house (please refer to the apparatus set up in Figure 1). A recording camera was positioned to face the arena, at an approximate 30° angle to the surface on which the arena is placed and 50 cm away from the transparent wall, so all mouse behaviour could be easily recorded for later analysis. The set-up also allows annotations to be made by independent experimenters to provide accurate and reproducible observations of the changes in behaviour. Pain behaviour was scored using a stopwatch when flinching, licking and flinching continued by licking, flicking or shaking of the paw were initiated and timer was stopped once behaviour ceased. Limping, altered locomotion or grooming of other parts of the body were not counted as pain behaviours. The first phase of the test was designated as the time between zero and 15 minutes after the formalin injection, whereas the late (or second) phase was from 15 minutes onwards (up to 45 minutes). Animals were humanely culled using Schedule 1 at the end of the test. All experiments were annotated by an experimenter, blinded to the genotype and treatment.\n\nStatistical analyses were performed using OriginLab 2017 software (Origin Group Corp.) (for area under the curve) and SPSS Statistics 20 (ANOVA repeated measures). For all sets of samples, normality tests were performed using the Shapiro-Wilk test, to check whether the data fitted in a Gaussian distribution (95% confidence intervals). Power calculations were performed using the Columbia University Biomath Calculator, following the guidelines previously described29. For details on the power calculation for each individual experiment, please refer to the Extended data section of this manuscript30. For all hypothesis testing, the minimum level of statistical significance adopted (p value) was at 0.05 - where if there was a 5% or less chance (5 in 100 or less) against the null hypothesis, so the latter would be rejected.\n\n\nResults\n\nThe formalin test is a robust model to study pain, and it has been demonstrated to be sensitive to various analgesic drugs10,31–34. The stress and anxiety-like behaviours triggered by aversive handling and restraining of the animal whilst being injected with the formalin, together with exposure to an unfamiliar environment, may interfere with the stimulus and the outcome of the test35. To check whether restraining stress could be minimised during the formalin injection, without having a major effect on the results of the test, we anaesthetised the animals with sevoflurane at the time of injection. Our results show that the anaesthetic virtually abolished the first phase of the test, whilst still preserving the second phase (Figure 2a)30. When compared to non-anaesthetised mice, we observed a reduction in pain behaviour of approximately 90% (Figure 2b) for anaesthetised mice during the first 15 minutes of the test (first phase). Notably, no significant changes in behaviour following the injection of anaesthetised mice were observed during the late phase of the test (20 – 45 min after injection) when compared to the non-anaesthetised group (Figure 2a & 2b).\n\n(a) Quantification of pain response over time of animals which were anaesthetised during the formalin injection (grey squares) and non-anaesthetised animals (white diamonds) shows distinct behaviors in the early phase of the test (0 - 15 min. after injection) but similar behaviours in the later phase (15 minutes onwards). ANOVA repeated measures: time: F(4,80): 5.92; p<0.001; timegroup: F(1,34): 5.357; p =0.001. (b) Plots representing the area under the curve show a significant reduction in pain response in anaesthetised vs. non-anaesthetised animals during the early phase of the test, but not during the late phase (p < 0.01; One-way ANOVA; N = 16 anaesthetised and N = 8 non-anaesthetised). Graphs represent means +/- SEM.\n\nThe effect of the anti-epileptic drug gabapentin as an analgesic is well documented. Indeed, many studies demonstrate that gabapentin attenuates nociceptive behaviour following formalin injection, specifically during the late phase of the test36–41. As the formalin test is widely used as a powerful tool to screen the analgesic effect of novel compounds at the preclinical level, we next tested whether gabapentin would also decrease the nociceptive behaviour observed after formalin injection using our newly proposed method. Our data demonstrate that in the gabapentin-treated group, there was a clear decrease in pain behaviour in the late phase of the test when compared to the control group (Figure 3a). Gabapentin treatment lead to a reduction of over 50% in nociceptive behaviour after formalin injection (Figure 3b) in comparison with the non-treated group.\n\nThe use of transgenic animals in research has been crucial in elucidating numerous biological mechanisms involved in both health and disease, including in the field of pain research. Given the importance of transgenic mice and the results of behavioural tests to screen potential molecular targets, we went on to investigate whether our refined formalin method could be effective for newly generated transgenic mouse lines. For these experiments, we used two knockout mouse lines in which the target genes are known to be expressed in the dorsal root ganglia (DRG)42–44 and in spinal cord neurons45.\n\nThe PTEN-induced kinase 1 (Pink1) gene has been extensively studied in the context of Parkinson’s disease46,47 and has been also linked to nociceptive processing48,49. Our data show that animals lacking Pink1 have a significantly lower response to the formalin test (Figure 4a). Although the Pink1 knockout group appeared to be marginally more responsive during the first 5 minutes of the test, their overall response to formalin injection during the first phase was very similar to their control littermates (Figure 4b). In contrast, their nociceptive behaviour during the late phase was reduced by 40% in comparison to the behaviour of the control group (Figure 4b). It appears that, despite being less responsive to the formalin, the Pink1 knockout animals seem to have a steadier response in the second phase when compared to the control littermates, with no obvious peak in the response over time (Figure 4a).\n\n(a) Graph showing the behavioural response of Pink1 -/- (pink squares) and their control littermates (white squares) over time. Pink1 -/- nociceptive behaviour in the late phase is reduced in comparison to the control group. ANOVA repeated measures, time: F(3.8, 41.5): 6.772; p< 0.001. (b) Area under the curve plot shows that whilst both groups have very similar response to formalin in the early phase of the test, Pink1 -/- mice display a lower pain response to formalin in the late phase (p < 0.001; One way ANOVA; N = 8 Pink1 -/- and N = 5 control). Graphs represent means +/- SEM.\n\nFollowing the same principle, we went on to screen a transgenic mouse line with a disrupted Slit Guidance Ligand 1 (Slit1) gene, using our refined formalin test. Slit1 is a secreted protein, which has been reported to be involved in DRG and spinal cord development, and previous studies have suggested a link between injury and increased Slit1 expression50–55. Our data show no overall difference in pain behaviour in the formalin test between the group with disrupted Slit1 function and their control littermates (Figure 5a). As with control mice, the pain behaviour during the first phase of the test is dramatically reduced due to the anaesthetic. Notably, a loss of function of the Slit1 gene did not lead to any change in the nociceptive behaviour during the late phase of the test (Figure 5a and 5b), as both groups spent a similar amount of time exhibiting pain behaviour (Figure 5b).\n\n(a) Graphs depicting the behavioural response to formalin injection in Slit1 -/- (blue triangles) mice and control littermates (white squares). Both groups show similar pain behaviour in both phases of the formalin test. ANOVA repeated measures, time: F(3.6, 72.9): 0.928; p< 0.001. (b) Area under the curve graph further shows the similarities in pain behaviour between Slit1 -/- and control groups. (n.s., where p> 0.05; One-way ANOVA; N = 11 Slit1 -/- and N = 11 control). Graphs represent means +/- SEM.\n\n\nDiscussion\n\nIn this study we proposed a modified and improved method for the formalin test. We showed that by anaesthetising the animal at the time of the injection reliable, robust and reproducible results are produced, whilst diminishing stress to the animals. Our newly proposed method showed that whilst the pain behaviour in the first phase was suppressed by the anaesthetic, the response in the late phase remained comparable to the control, non-anaesthetised group. Our results were validated by pharmacological suppression of the late phase with gabapentin. Furthermore, we demonstrate that this is a useful method to use while screening transgenic lines for changes in pain behaviour in response to formalin.\n\nReducing the levels of stress in the animals is of great advantage when performing behaviour tests. Many studies have shown that mishandling of laboratory animals can have profound impacts on some behavioural tests35,56,57. In particular, animal restraining has a drastic, negative effect on anxiety levels58,59, with evidence suggesting that the adverse effects also extend to pain behaviour, where it can lead to areflexia, hyperalgesia and, in some cases, abnormal flinching60–64. Given the potential adverse effect that stress, resulting from either mishandling or restraining the animal during the administration of a drug or compounds, can have on pain behaviour, we propose the use of anaesthesia during the drug application. In our method, we show that animals can be injected with formalin without causing any apparent distress, whilst increasing the reproducibility of the test, as consistency in the site of injection as well as the volume injected is improved under anaesthesia. Indeed, due to ethical considerations, it has recently been suggested that brief anaesthesia immediately before injection would be beneficial for the animals, as well as increasing the accuracy of substance application, considering that the site of formalin injection is crucial65. Importantly, we show that the outcome of the test is similar to that of the traditional method and that, despite the anaesthesia reducing the pain behaviour during first phase, the second phase behaviour times remained almost identical. Notably, our method has a significant advantage over a previous study which employed different inhalation anaesthetics, shown to have a negative impact on the early and late stages of the test66, suppressing the pain behaviour in both phases dramatically and being remarkably different to non-anaesthetised animals. In summary, our method resulted in minor changes to overall behaviour responses and provided significant advantages for ethical and stress-free animal handling.\n\nCentral sensitisation by formalin appears to be the most crucial aspect of the test when evaluating nocifensive behaviour. The underlying mechanisms of the formalin test are not fully understood. Historical experimental data indicate that the behavioural response observed after the injection is solely due to the direct stimulation and activity of C-fibre nociceptors1,9,67, whereas subsequent studies suggest the involvement of Aδ and non-nociceptive Aβ-fibres68,69. Whilst there is still controversy regarding the circuitry and cellular and molecular mechanisms triggered by formalin, subjects which are beyond the scope of this study, we demonstrate that the second phase of the test can be used to screen pain behaviour independently of the first phase. Supporting our findings, studies showed that knockout or ablation of distinct nerve fibre populations in animal models resulted mostly in a reduction of the pain behaviour in the second phase of the test, while the first phase was not necessarily affected69–71. Furthermore, these studies show that, in all instances in which the first phase is affected, the second phase will also be influenced69–71, demonstrating therefore that suppressing or diminishing the pain behaviour during the first phase of the test is almost inconsequential when screening phenotypes and testing drugs. Therefore, our study highlights that the formalin test could be improved by diminishing unnecessary animal distress without compromising the results, given that the first phase of the test is, in most cases, not very informative.\n\nThe second phase of the formalin test alone can be used for phenotypic screening. We confirmed the sensitivity of our modified protocol by showing that the effects of the extensively characterised analgesic, gabapentin36–40 are maintained as expected. We further validated the sensitivity of the modified formalin procedure in two different transgenic lines. The mitochondrial kinase Pink1 has been extensively studied, and mutations in this gene are notoriously linked to neuronal dysfunction in Parkinson’s disease47,72–74. Previous studies also linked the loss of function of Pink1 in humans with abnormal pain sensation, where subjects present a higher mechanical and pressure threshold48,49. Notably, and similarly to the phenotype found in humans, we show here that animals with a disrupted Pink1 gene display lower nociceptive behaviour after formalin injection. However, it cannot be excluded, the hypoalgesic phenotype can be due to symptoms arising from Parkinson’s disease itself - as Pink1-null mouse models might present motor dysfunction75. Therefore, our results not only present for the first time a pain phenotype in a Pink1-null rodent model in the context of the formalin test, but also supports the refinement of the proposed formalin method. Further to the pain phenotype observed in this Pink1 knockout line, we also screened for any distinct phenotype observed for the Slit1 knockout mice. Despite previous studies suggesting a role for Slit1 in neuronal development51,52,76, our data shows that global deletion of the gene does not lead to any alteration in pain behaviour in the formalin test. Together, these results demonstrate that the refined formalin test proposed in this study can be broadly used, not only to test the efficacy of drugs, as shown with gabapentin, but also to evaluate pain phenotypes in newly generated transgenic models.\n\nIn conclusion, in this study we present a refinement to the already established formalin test. We propose that the use of an inhalable anaesthetic during the injection of formalin is not only a reliable method to improve consistency when injecting the compound, but most importantly, represents a valuable refinement. We show that this method complies with the 3Rs sought by ethical committees, as well as meeting the additional 3Rs (relevancy, robustness and repeatability) sought by scientists13. Moreover, we demonstrate that the test is sensitive enough to screen for possible pain phenotypes and suggest that diminishing the first phase of the formalin test has little consequence on the global pain response of the animal.\n\n\nData availability\n\nFigshare: Lopes et al., A refinement to the formalin test in mice - Extended data. https://doi.org/10.6084/m9.figshare.8230655.v130\n\nThis project contains the following extended data:\n\n- Data_Fig_2.csv (raw data underlying Figure 2)\n\n- Data_Fig_3.csv (raw data underlying Figure 3)\n\n- Data_Fig_4.csv (raw data underlying Figure 4)\n\n- Data_Fig_5.csv (raw data underlying Figure 5)\n\nFigshare: Lopes et al., A refinement to the formalin test in mice - Extended data. https://doi.org/10.6084/m9.figshare.8230655.v130\n\nThis project contains the following extended data:\n\n- Lopes_el_al.,_2019_Power_Calculations.pdf (details of power calculations for each experiment)\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).", "appendix": "Grant information\n\nThis work was supported by the Wellcome Trust [102645]; Medical Research Council [53658, 53650].\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe would like to thank Dr Franziska Denk for useful discussions of the data and advice on statistical analysis. We thank Dr Kim Chisholm and Dr Michelle Edye for the comments on the manuscript. This work was supported by the Wellcome Trust Program (102645/Z/13/Z). Mice were obtained from MRC Harwell with funding from the Medical Research Council for generating (53658) and phenotyping (53650), a member of the International Phenotyping Consortium (funding and further phenotypic information can be found at http://www.mousephenotype.org).\n\n\nReferences\n\nDubuisson D, Dennis SG: The formalin test: a quantitative study of the analgesic effects of morphine, meperidine, and brain stem stimulation in rats and cats. Pain. 1977; 4(2): 161–74. PubMed Abstract \| Publisher Full Text\n\nCoderre TJ, Melzack R: The role of NMDA receptor-operated calcium channels in persistent nociception after formalin-induced tissue injury. J Neurosci. 1992; 12(9): 3671–5. PubMed Abstract \| Publisher Full Text\n\nErami E, Azhdari-Zarmehri H, Imoto K, et al.: Characterization of Nociceptive Behaviors Induced by Formalin in the Glabrous and Hairy Skin of Rats. Basic Clin Neurosci. 2017; 8(1): 37–42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAbbott FV, Franklin KB, Ludwick RJ, et al.: Apparent lack of tolerance in the formalin test suggests different mechanisms for morphine analgesia in different types of pain. Pharmacol Biochem Behav. 1981; 15(4): 637–40. PubMed Abstract \| Publisher Full Text\n\nRosland JH, Tjolsen A, Maehle B, et al.: The formalin test in mice: effect of formalin concentration. Pain. 1990; 42(2): 235–42. PubMed Abstract \| Publisher Full Text\n\nSawynok J, Liu XJ: The Formalin test: characteristics and usefulness of the model. Reviews in Analgesia. 2004; 7(2): 145–163. Publisher Full Text\n\nTjolsen A, Berge OG, Hunskaar S, et al.: The formalin test: an evaluation of the method. Pain. 1992; 51(1): 5–17. PubMed Abstract \| Publisher Full Text\n\nHole K, Tjolsen A: The tail-flick and formalin tests in rodents: changes in skin temperature as a confounding factor. Pain. 1993; 53(3): 247–54. PubMed Abstract \| Publisher Full Text\n\nHunskaar S, Fasmer OB, Hole K: Formalin test in mice, a useful technique for evaluating mild analgesics. J Neurosci Methods. 1985; 14(1): 69–76. PubMed Abstract \| Publisher Full Text\n\nHunskaar S, Hole K: The formalin test in mice: dissociation between inflammatory and non-inflammatory pain. Pain. 1987; 30(1): 103–14. PubMed Abstract \| Publisher Full Text\n\nShibata M, Ohkubo T, Takahashi H, et al.: Modified formalin test: characteristic biphasic pain response. Pain. 1989; 38(3): 347–52. PubMed Abstract \| Publisher Full Text\n\nMurray CW, Porreca F, Cowan A: Methodological refinements to the mouse paw formalin test. An animal model of tonic pain. J Pharmacol Methods. 1988; 20(2): 175–86. PubMed Abstract \| Publisher Full Text\n\nWells DJ: Animal welfare and the 3Rs in European biomedical research. Ann N Y Acad Sci. 2011; 1245: 14–6. PubMed Abstract \| Publisher Full Text\n\nUvarov O: Research with animals: requirement, responsibility, welfare. Lab Anim. 1985; 19(1): 51–75. PubMed Abstract \| Publisher Full Text\n\nParker RM, Browne WJ: The place of experimental design and statistics in the 3Rs. ILAR J. 2014; 55(3): 477–85. PubMed Abstract \| Publisher Full Text\n\nRussell WMS, Burch RL: The principles of humane experimental technique. London: Methuen & Co., 1959. Reference Source\n\nSneddon LU, Halsey LG, Bury NR: Considering aspects of the 3Rs principles within experimental animal biology. J Exp Biol. 2017; 220(Pt 17): 3007–3016. PubMed Abstract \| Publisher Full Text\n\nResch M, Neels T, Tichy A, et al.: Impact assessment of tail-vein injection in mice using a modified anaesthesia induction chamber versus a common restrainer without anaesthesia. Lab Anim. 2019: 53(2): 190–201. PubMed Abstract \| Publisher Full Text\n\nKim JW, Ko MJ, Gonzales EL, et al.: Social support rescues acute stress-induced cognitive impairments by modulating ERK1/2 phosphorylation in adolescent mice. Sci Rep. 2018; 8(1): 12003. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSutanto W, de Kloet ER: The use of various animal models in the study of stress and stress-related phenomena. Lab Anim. 1994; 28(4): 293–306. PubMed Abstract \| Publisher Full Text\n\nKovacs LA, Schiessl JA, Nafz AE, et al.: Both Basal and Acute Restraint Stress-Induced c-Fos Expression Is Influenced by Age in the Extended Amygdala and Brainstem Stress Centers in Male Rats. Front Aging Neurosci. 2018; 10: 248. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchaefer DC, Asner IN, Seifert B, et al.: Analysis of physiological and behavioural parameters in mice after toe clipping as newborns. Lab Anim. 2010; 44(1): 7–13. PubMed Abstract \| Publisher Full Text\n\nNohara M, Tohei A, Sato T, et al.: Evaluation of response to restraint stress by salivary corticosterone levels in adult male mice. J Vet Med Sci. 2016; 78(5): 775–80. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMeijer MK, Spruijt BM, van Zutphen LF, et al.: Effect of restraint and injection methods on heart rate and body temperature in mice. Lab Anim. 2006; 40(4): 382–91. PubMed Abstract \| Publisher Full Text\n\nPoole T: Happy animals make good science. Lab Anim. 1997; 31(2): 116–24. PubMed Abstract \| Publisher Full Text\n\nJennings EM, Okine BN, Roche M, et al.: Stress-induced hyperalgesia. Prog Neurobiol. 2014; 121: 1–18. PubMed Abstract \| Publisher Full Text\n\nPorro CA, Carli G: Immobilization and restraint effects on pain reactions in animals. Pain. 1988; 32(3): 289–307. PubMed Abstract \| Publisher Full Text\n\nImbe H, Iwai-Liao Y, Senba E: Stress-induced hyperalgesia: animal models and putative mechanisms. Front Biosci. 2006; 11(1): 2179–92. PubMed Abstract \| Publisher Full Text\n\nCharan J, Kantharia ND: How to calculate sample size in animal studies? J Pharmacol Pharmacother. 2013; 4(4): 303–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLopes D, Cater H, Thakur M, et al.: Lopes et al., A refinement to the formalin test in mice - Extended data. 2019. http://www.doi.org/10.6084/m9.figshare.8230655.v1\n\nAfify EA, Alkreathy HM, Ali AS, et al.: Characterization of the Antinociceptive Mechanisms of Khat Extract (Catha edulis) in Mice. Front Neurol. 2017; 8: 69. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nClarke PB, Franklin KB: Infusions of 6-hydroxydopamine into the nucleus accumbens abolish the analgesic effect of amphetamine but not of morphine in the formalin test. Brain Res. 1992; 580(1–2): 106–10. PubMed Abstract \| Publisher Full Text\n\nDamas J, Liegeois JF: The inflammatory reaction induced by formalin in the rat paw. Naunyn Schmiedebergs Arch Pharmacol. 1999; 359(3): 220–7. PubMed Abstract \| Publisher Full Text\n\nTaylor BK, Peterson MA, Roderick RE, et al.: Opioid inhibition of formalin-induced changes in plasma extravasation and local blood flow in rats. Pain. 2000; 84(2–3): 263–70. PubMed Abstract \| Publisher Full Text\n\nGouveia K, Hurst JL: Optimising reliability of mouse performance in behavioural testing: the major role of non-aversive handling. Sci Rep. 2017; 7: 44999. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCarlton SM, Zhou S: Attenuation of formalin-induced nociceptive behaviors following local peripheral injection of gabapentin. Pain. 1998; 76(1–2): 201–7. PubMed Abstract \| Publisher Full Text\n\nField MJ, Oles RJ, Lewis AS, et al.: Gabapentin (neurontin) and S-(+)-3-isobutylgaba represent a novel class of selective antihyperalgesic agents. Br J Pharmacol. 1997; 121(8): 1513–22. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLaughlin TM, Tram KV, Wilcox GL, et al.: Comparison of antiepileptic drugs tiagabine, lamotrigine, and gabapentin in mouse models of acute, prolonged, and chronic nociception. J Pharmacol Exp Ther. 2002; 302(3): 1168–75. PubMed Abstract \| Publisher Full Text\n\nMunro G: Pharmacological assessment of the rat formalin test utilizing the clinically used analgesic drugs gabapentin, lamotrigine, morphine, duloxetine, tramadol and ibuprofen: influence of low and high formalin concentrations. Eur J Pharmacol. 2009; 605(1–3): 95–102. PubMed Abstract \| Publisher Full Text\n\nShimoyama N, Shimoyama M, Davis AM, et al.: Spinal gabapentin is antinociceptive in the rat formalin test. Neurosci Lett. 1997; 222(1): 65–67. PubMed Abstract \| Publisher Full Text\n\nSalinas-Abarca AB, Avila-Rojas SH, Barragan-Iglesias P, et al.: Formalin injection produces long-lasting hypersensitivity with characteristics of neuropathic pain. Eur J Pharmacol. 2017; 797: 83–93. PubMed Abstract \| Publisher Full Text\n\nFlegel C, Schöbel N, Altmuller J, et al.: RNA-Seq Analysis of Human Trigeminal and Dorsal Root Ganglia with a Focus on Chemoreceptors. PLoS One. 2015; 10(6): e0128951. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThakur M, Crow M, Richards N, et al.: Defining the nociceptor transcriptome. Front Mol Neurosci. 2014; 7: 87. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nUsoskin D, Furlan A, Islam S, et al.: Unbiased classification of sensory neuron types by large-scale single-cell RNA sequencing. Nat Neurosci. 2015; 18(1): 145–153. PubMed Abstract \| Publisher Full Text\n\nSakurai M, Kawamura T, Nishimura H, et al.: Induction of Parkinson disease-related proteins in motor neurons after transient spinal cord ischemia in rabbits. J Cereb Blood Flow Metab. 2009; 29(4): 752–758. PubMed Abstract \| Publisher Full Text\n\nKawajiri S, Saiki S, Sato S, et al.: Genetic mutations and functions of PINK1. Trends Pharmacol Sci. 2011; 32(10): 573–580. PubMed Abstract \| Publisher Full Text\n\nPickrell AM, Youle RJ: The roles of PINK1, parkin, and mitochondrial fidelity in Parkinson's disease. Neuron. 2015; 85(2): 257–273. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGierthmuhlen J, Lienau F, Maag R, et al.: Somatosensory processing in a German family with PINK1 mutations: its potential role in Parkinson disease. J Neurol Neurosurg Psychiatry. 2009; 80(5): 571–574. PubMed Abstract \| Publisher Full Text\n\nLudwig J, Lienau F, Maag R, et al.: Abnormal somatosensory processing in PINK1 (PARK6) mutations carriers. Eur J Pain. 2006; 10(S1): S135. Publisher Full Text\n\nCarr L, Parkinson DB, Dun XP: Expression patterns of Slit and Robo family members in adult mouse spinal cord and peripheral nervous system. PLoS One. 2017; 12(2): e0172736. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLong H, Sabatier C, Ma L, et al.: Conserved roles for Slit and Robo proteins in midline commissural axon guidance. Neuron. 2004; 42(2): 213–223. PubMed Abstract \| Publisher Full Text\n\nMambetisaeva ET, Andrews W, Camurri L, et al.: Robo family of proteins exhibit differential expression in mouse spinal cord and Robo-Slit interaction is required for midline crossing in vertebrate spinal cord. Dev Dyn. 2005; 233(1): 41–51. PubMed Abstract \| Publisher Full Text\n\nWehrle R, Camand E, Chedotal A, et al.: Expression of netrin-1, slit-1 and slit-3 but not of slit-2 after cerebellar and spinal cord lesions. Eur J Neurosci. 2005; 22(9): 2134–2144. PubMed Abstract \| Publisher Full Text\n\nYi XN, Zheng LF, Zhang JW, et al.: Dynamic changes in Robo2 and Slit1 expression in adult rat dorsal root ganglion and sciatic nerve after peripheral and central axonal injury. Neurosci Res. 2006; 56(3): 314–321. PubMed Abstract \| Publisher Full Text\n\nZhang HY, Zheng LF, Yi XN, et al.: Slit1 promotes regenerative neurite outgrowth of adult dorsal root ganglion neurons in vitro via binding to the Robo receptor. J Chem Neuroanat. 2010; 39(4): 256–261. PubMed Abstract \| Publisher Full Text\n\nDeacon RM: Housing, husbandry and handling of rodents for behavioral experiments. Nat Protoc. 2006; 1(2): 936–946. PubMed Abstract \| Publisher Full Text\n\nHurst JL, West RS: Taming anxiety in laboratory mice. Nat Methods. 2010; 7(10): 825–826. PubMed Abstract \| Publisher Full Text\n\nBuynitsky T, Mostofsky DI: Restraint stress in biobehavioral research: Recent developments. Neurosci Biobehav Rev. 2009; 33(7): 1089–1098. PubMed Abstract \| Publisher Full Text\n\nChiba S, Numakawa T, Ninomiya M, et al.: Chronic restraint stress causes anxiety- and depression-like behaviors, downregulates glucocorticoid receptor expression, and attenuates glutamate release induced by brain-derived neurotrophic factor in the prefrontal cortex. Prog Neuropsychopharmacol Biol Psychiatry. 2012; 39(1): 112–119. PubMed Abstract \| Publisher Full Text\n\nFuchs PN, Melzack R: Restraint reduces formalin-test pain but the effect is not influenced by lesions of the hypothalamic paraventricular nucleus. Exp Neurol. 1996; 139(2): 299–305. PubMed Abstract \| Publisher Full Text\n\nKing CD, Devine DP, Vierck CJ, et al.: Opioid modulation of reflex versus operant responses following stress in the rat. Neuroscience. 2007; 147(1): 174–182. PubMed Abstract \| Publisher Full Text\n\nKing CD, Devine DP, Vierck CJ, et al.: Differential effects of stress on escape and reflex responses to nociceptive thermal stimuli in the rat. Brain Res. 2003; 987(2): 214–222. PubMed Abstract \| Publisher Full Text\n\nGameiro GH, Gameiro PH, Andrade Ada S, et al.: Nociception- and anxiety-like behavior in rats submitted to different periods of restraint stress. Physiol Behav. 2006; 87(4): 643–649. PubMed Abstract \| Publisher Full Text\n\nTaylor BK, Peterson MA, Basbaum AI: Exaggerated cardiovascular and behavioral nociceptive responses to subcutaneous formalin in the spontaneously hypertensive rat. Neurosci Lett. 1995; 201(1): 9–12. PubMed Abstract \| Publisher Full Text\n\nCapone F, Aloisi AM: Refinement of pain evaluation techniques. The formalin test. Ann Ist Super Sanita. 2004; 40(2): 223–229. PubMed Abstract\n\nFukuda T, Nishimoto C, Hisano S, et al.: The analgesic effect of xenon on the formalin test in rats: a comparison with nitrous oxide. Anesth Analg. 2002; 95(5): 1300–1304, table of contents. PubMed Abstract\n\nRussell N N, Heapy CG, Jamieson A: Afferent activity in models of inflammation. Pain. 1987; 30: S255. Publisher Full Text\n\nPuig S, Sorkin LS: Formalin-evoked activity in identified primary afferent fibers: systemic lidocaine suppresses phase-2 activity. Pain. 1996; 64(2): 345–355. PubMed Abstract \| Publisher Full Text\n\nShields SD, Cavanaugh DJ, Lee H, et al.: Pain behavior in the formalin test persists after ablation of the great majority of C-fiber nociceptors. Pain. 2010; 151(2): 422–429. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMacpherson LJ, Xiao B, Kwan KY, et al.: An ion channel essential for sensing chemical damage. J Neurosci. 2007; 27(42): 11412–11415. PubMed Abstract \| Publisher Full Text\n\nMcNamara CR, Mandel-Brehm J, Bautista DM, et al.: TRPA1 mediates formalin-induced pain. Proc Natl Acad Sci U S A. 2007; 104(33): 13525–13530. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMouton-Liger F, Jacoupy M, Corvol JC, et al.: PINK1/Parkin-Dependent Mitochondrial Surveillance: From Pleiotropy to Parkinson's Disease. Front Mol Neurosci. 2017; 10: 120. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPrzedborski S: The two-century journey of Parkinson disease research. Nat Rev Neurosci. 2017; 18(4): 251–259. PubMed Abstract \| Publisher Full Text\n\nScarffe LA, Stevens DA, Dawson VL, et al.: Parkin and PINK1: much more than mitophagy. Trends Neurosci. 2014; 37(6): 315–324. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGlasl L, Kloos K, Giesert F, et al.: Pink1-deficiency in mice impairs gait, olfaction and serotonergic innervation of the olfactory bulb. Exp Neurol. 2012; 235(1): 214–227. PubMed Abstract \| Publisher Full Text\n\nMarin O, Plump AS, Flames N, et al.: Directional guidance of interneuron migration to the cerebral cortex relies on subcortical Slit1/2-independent repulsion and cortical attraction. Development. 2003; 130(9): 1889–1901. PubMed Abstract \| Publisher Full Text" }	[ { "id": "50153", "date": "26 Jun 2019", "name": "Nora Bourbia", "expertise": [ "Reviewer Expertise I have done a PhD in neurophysiology of pain using rat model of chronic pain", "followed by a postdoc in neurobehavioral genetics using mouse models. I am now a newPI in neurobiology." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nDouglas M Lopes et al. have addressed the question of whether it is possible to refine the formalin test without affecting the behavioural outcome important in pain research. They have shown that the use of anaesthesia (sevoflurane) at the injection time reduces the distress of the handling and formalin injection, and the behaviour associated to the first phase while conserving the behaviour associated to the late phase of the formalin injection test.\nThe experiment and the method are well performed, clear to follow for reproducibility and well designed to answer to the initial question. The study mentioned that both female and male mice were used which is very valuable and important in research. However, and except if I have missed it, I am not aware of the presence of a statistical test performed to confirm the absence of a sex-difference in behavioural responses to the formalin test in this study because sex difference can be a factor influencing the pain perception. A clarification would be appreciated as well as the presence of the sex in the source data. Additionally, it is mentioned that wildtype C57BL/6 mice were used for the gabapentin and sevoflurane experiments but without mentioning the exact background while it is clearly mentioned for the transgenic mouse study (C57BL/6NTac). Pain-like behaviour can be influenced by the mouse strains, also behavioural differences have been observed between C57BL/6N and C57BL/6J mice. It would be valuable to mention the exact background for a full reproducibility.\nThe results of the sevoflurane effect, the gabapentin treatment, and Slit1 KO studies are very clear. I did, however, notice a discrepancy in the data from the control group of the Pink1-/- study compared to the control group in Slit-/- study (both C57BL/6Ntac background) and the wildtype C57BL/6 mice in the anaesthesia/gabapentin experiment. The average pain behaviour in the control mice of the Pink1-/- study is almost twice higher than all other group controls and the Pink1-/- mice between the time 20 and 35 min. Could the authors provide a possible explanation about this discrepancy? This might lead to a potential false positive result of Pink1 effect in this study. Ideally, to confirm the effect of Pink1-/- on the behaviour induced by the formalin test, it would be appreciated to repeat this test (Pink1-/- vs its littermate control) with again randomisation and blinding.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly", "responses": [ { "c_id": "4787", "date": "02 Aug 2019", "name": "Douglas Lopes", "role": "Author Response", "response": "We would like to thank the reviewers for the very positive reviews of our paper. The reviewer’s comments were very helpful, and we have now revised our manuscript to address all of the issues raised. The study mentioned that both female and male mice were used which is very valuable and important in research. However, and except if I have missed it, I am not aware of the presence of a statistical test performed to confirm the absence of a sex-difference in behavioural responses to the formalin test in this study because sex difference can be a factor influencing the pain perception. A clarification would be appreciated as well as the presence of the sex in the source data. Thank you for pointing that out. We have now uploaded the files containing the required information. We have also conducted statistical tests comparing males and females across the different experiments (except for the experiments using gabapentin, where only males were used). Furthermore, we uploaded an additional figure where the graphs contain each individual animal (females in red circle and males in blue square). The figure legend contains the details for each one of the tests. Both the visual representations and statistical analysis demonstrate no differences between the genders. A sentence stating the above findings and link to the Supplementary Figure has been added to the Material and methods section (Experimental Animals). It should be noted that although we performed the appropriate statistical test to compare sex-differences and observe no differences, for some of the groups analysed the outcome might be pursued as not ‘scientifically meaningful’, as the numbers of animals in each subgroup were below the ‘n’ number recommended by some recent published guidelines (Curtis, et al., 2018 Br J Pharmacol. Apr; 175(7): 987–993). Additionally, it is mentioned that wildtype C57BL/6 mice were used for the gabapentin and sevoflurane experiments but without mentioning the exact background while it is clearly mentioned for the transgenic mouse study (C57BL/6NTac). Pain-like behaviour can be influenced by the mouse strains, also behavioural differences have been observed between C57BL/6N and C57BL/6J mice. It would be valuable to mention the exact background for a full reproducibility. The background has now been added (C57BL/6NTac). The results of the sevoflurane effect, the gabapentin treatment, and Slit1 KO studies are very clear. I did, however, notice a discrepancy in the data from the control group of the Pink1-/- study compared to the control group in Slit-/- study (both C57BL/6Ntac background) and the wildtype C57BL/6 mice in the anaesthesia/gabapentin experiment. The average pain behaviour in the control mice of the Pink1-/- study is almost twice higher than all other group controls and the Pink1-/- mice between the time 20 and 35 min. Could the authors provide a possible explanation about this discrepancy? This might lead to a potential false positive result of Pink1 effect in this study. Ideally, to confirm the effect of Pink1-/- on the behaviour induced by the formalin test, it would be appreciated to repeat this test (Pink1-/- vs its littermate control) with again randomisation and blinding. Thank you for pointing this out. We trust the behaviour observed represents the natural variation it can be obtained when performing behavioural tests and therefore we can only emphasise why it is important to control with animals that have been bred at the same time. Although we understand and acknowledge the reviewer’s concerns, we trust the data is a true representation of the mice behaviour for this particular line. As demonstrated on the graph (Supplementary Figure 1 B), all the control animals used in this test were above the average control response in the other figures (above ~250 sec.), we therefore believe it is unlikely the findings are a false positive. We hypothesise that there is not a biological obvious reason that can explain this variation apart from the fact that they are a different batch of animals and thus likely to differ in experimenter, day and uncontrollable variances in their cage environment (such as genotype of parents, litters, litter size, etc.). Unfortunately, this line is no longer promptly available (it has been cryopreserved and archived) and therefore we are regrettably unable to repeat these experiments to investigate this further. We have however, added a sentence in the manuscript pointing out the differences observed and explaining it as above." } ] }, { "id": "50152", "date": "22 Jul 2019", "name": "John Riddell", "expertise": [ "Reviewer Expertise Animal models of pain", "behavioral testing of pain-like behaviors", "electrophysiology", "dorsal horn circuit" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this article, the authors demonstrate a refinement of the widely-used formalin test. They show that if the formalin injection is performed while mice are under brief general anaesthesia (Sevoflurane) the second phase behaviour is largely unaltered, even though the first phase behaviour is largely lost. Since the second phase is the more useful measure, this means that the procedure can be performed in a less stressful way, resulting in a significant improvement in animal welfare. The results of the study are generally convincing, and the article is clearly written and well illustrated. One concern is the relatively high level of pain behaviour in the control group for the Pink1 study (Fig 4). While the area under the curve for other groups is typically around 1000, for the controls in Fig 4 it is nearly 2000. In fact, compared to other cohorts, the Pink1-/- mice seem to have normal behaviour, while the control group show an exaggerated response. Some explanation is needed here. Also – what are the units for area under curve?\nMinor points:\n\nIt would be helpful to have some explanation of why the Pink1 and Slit1 mouse lines were chosen for this study, at the end of the Experimental Animals paragraph in the Methods section.\n\nPage 4: In the 4th line of the Results section \"it is\" should be inserted before \"being\" – otherwise it would be the experimenter who received the injection.\n\nPage 5: left column, 6th line \"led\" rather than \"lead\".\n\nPage 6: left column, 6th line – omit \"seem to\".\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly", "responses": [ { "c_id": "4788", "date": "02 Aug 2019", "name": "Douglas Lopes", "role": "Author Response", "response": "We would like to thank the reviewers for the very positive reviews of our paper. The reviewer’s comments were very helpful, and we have now revised our manuscript to address all of the issues raised. The results of the study are generally convincing, and the article is clearly written and well illustrated. One concern is the relatively high level of pain behaviour in the control group for the Pink1 study (Fig 4). While the area under the curve for other groups is typically around 1000, for the controls in Fig 4 it is nearly 2000. In fact, compared to other cohorts, the Pink1-/- mice seem to have normal behaviour, while the control group show an exaggerated response. Some explanation is needed here. This point was was also raised by the other reviewer and addressed accordingly. We trust the behaviour observed represents the natural variation it can be obtained when performed behavioural tests and therefore we can only emphasise why it is important to control with animals that have been bred at the same time. Although we understand and acknowledge the reviewer’s concerns, we trust the data is a true representation of the mice behaviour for this particular line. As demonstrated on the graph (Supplementary Figure 1 B), all the control animals used in this test were above the average control response in the other figures (above ~250 sec.), we therefore believe it is unlikely the findings are a false positive. We hypothesise that there is not a biological obvious reason that can explain this variation apart from the fact that they are a different batch of animals and thus likely to differ in experimenter, day and uncontrollable variances in their cage environment (such as genotype of parents, litters, litter size, etc.). Unfortunately, this line is no longer promptly available (it has been cryopreserved and archived) and therefore we are regrettably unable to repeat these experiments to investigate this further. We have however, added a sentence in the manuscript pointing out the differences observed and explaining it as above. Also – what are the units for area under curve? The AUC was calculated in relation to the pain response (sec) over time. This has been added to the materials and methods Minor points: It would be helpful to have some explanation of why the Pink1 and Slit1 mouse lines were chosen for this study, at the end of the Experimental Animals paragraph in the Methods section. These sentences were added: The transgenic lines chosen in this study were part of a parallel neuroscience program study being carried out at the MLC. In addition, it has been suggested a link between Pink1 and nociceptive processing (48,49) and Slit1 expression and peripheral injury (50–55). Page 4: In the 4th line of the Results section \"it is\" should be inserted before \"being\" – otherwise it would be the experimenter who received the injection. This has been amended. Page 5: left column, 6th line \"led\" rather than \"lead\". This has been amended. Page 6: left column, 6th line – omit \"seem to\". This has been amended." } ] } ]	1	https://f1000research.com/articles/8-891
https://f1000research.com/articles/8-1340/v1	02 Aug 19	{ "type": "Research Article", "title": "Whole-genome sequence analysis of Vibrio cholerae from three outbreaks in Uganda, 2014 - 2016", "authors": [ "Dickson Aruhomukama", "Ivan Sserwadda", "Gerald Mboowa", "Ivan Sserwadda" ], "abstract": "Background: Cholera remains a serious public health problem in Uganda and Africa. The aim of this study was to provide the complete array of antimicrobial resistance genes, integrative and conjugative elements, virulence genes, pathogenicity islands, plasmids, and insertion sequences in the strains. In addition, this study also aimed to provide a single nucleotide polymorphism (SNP) based phylogenetic analysis of the strains. Methods: In the analysis, both Linux and web-based bioinformatics approaches were used to analyze the study sequences. Databases used included; FastQC, MultiQC, Snippy, PANTHER, PATRIC, Unicycler, ISFinder, Center for Genomic Epidemiology pipelines (i.e. MLST, PlasmidFinder, MyDbFinder, and ResFinder), MashTree and IcyTree. Results: The 10 sequenced strains of Vibrio cholerae were found to carry virulence-associated genes including MakA, ctxA, ctxB, carA, carB, trpB, clpB, ace, toxR, zot, rtxA, ompW, ompR, gmhA, fur, hlyA, and rstR. Also identified were: genes of the Type VI secretion system including vasA-L, vgrG-2, vgrG-3, vipA/mglA, and vipB/mglB; alsD (VC1589), involved in the synthesis of 2,3-butanediol; alsR, involved in the acetate-responsive LysR-type regulation; makA, the flagella-mediated cytotoxin gene; Type VI pilus genes including tcpA-F, tcpH-J, tcpN, tcpP-T, and icmF/vasK; adherence genes acfA-D and IlpA; and quorum sensing system genes luxS and cqsA. Pathogenicity islands identified comprised of VSP-1 and VSP-2, as well as VPI-1 and VPI-2. In addition, strA and B, APH(3'')-I, APH(3'')-Ib, APH(6)-Id, APH(6)-Ic, murA, pare, dfrA1, floR, catB, and catB9 were among the antimicrobial resistance genes found in the sequences. Analysis for SNPs shared among the sequences showed that the sequenced strains shared 218 SNPs and of these, 98 SNPs were missense. Gene enrichment analysis of these SNPs showed enrichment in genes that mediate transmembrane-signaling receptor activity, peptidyl-prolyl cis-trans isomerase activity, and phosphor-relay response regulator activity. Conclusions: This study applied bioinformatics approaches to provide comprehensive genomic analysis of V. cholerae genomes obtained from Uganda.", "keywords": [ "Vibrio cholerae", "Whole-genome sequencing", "Bioinformatics", "Genomics" ], "content": "Introduction\n\nCholera remains a serious public health problem in Uganda and Africa as a whole1,2. It is characterized by a large disease burden, recurrent outbreaks, high case fatality rates, as well as tenacious endemicity1,2. Over the last four decades, Uganda has experienced several cholera outbreaks2. The detection, monitoring, and surveillance of cholera in Uganda rely upon the isolation of Vibrio cholerae using culture-based methods in microbiology laboratories2,3. However, these methods are faced with several challenges, including: associated long turn-around times (24–48 hours); limited microbiology laboratory capacity; lack of laboratory supplies; poor laboratory infrastructure, particularly electricity necessary to operate laboratory equipment; as well as limited reliable and rapid diagnostic tests2. Unlike culture-based methods, high-throughput sequencing, a culture-independent method, has been documented to be less affected by most of the challenges facing culture-based methods and at the same time provides an unprecedented view of pathogen biology and delivers high-resolution genomic epidemiology via rapid and cheap whole-genome sequencing4–7. Despite this knowledge, sequencing remains a less desirable option for most scientists in Uganda and hence, data on the genomic characteristics of V. cholerae remains scarce, likely attributable to the underdeveloped bioinformatics capacity and lack of expertise necessary for analyzing whole-genome sequencing data7. This study set out to use bioinformatics approaches to analyze whole-genome sequence data obtained from V. cholerae isolates from different outbreaks in Uganda, with the aim of providing the complete array of virulence genes, pathogenicity islands, antimicrobial resistance genes, integrative and conjugative elements, and antimicrobial resistance genes associated with these elements, plasmids, and insertion sequences. In addition, this study also provided a single nucleotide polymorphism (SNP) based phylogenetic analysis of the strains.\n\n\nMethods\n\nThis was a cross-sectional study that analyzed 10 whole-genome sequences of V. cholerae isolates. These isolates were collected during three different cholera outbreaks in Uganda between 2014 and 2016 and sequenced by a group from the University of Maryland (Bwire et al., 2018). The whole-genome sequencing data was deposited in the NCBI’s Sequence Read Archive (SRA) with the accession number SRP136117.\n\nProcedures and considerations in sample collection and whole-genome sequencing are described by Bwire et al., 2018. Briefly, whole-genome sequencing was done using three or four representative samples that have been obtained from each of the three Multiple-Locus Variable Number Tandem-Repeat Analysis clonal complexes that had been identified during the period 2014–2016. Steps in whole-genome sequencing were: library preparation from fragmented DNA, this was achieved using an appropriate library preparation kit (KAPA High Throughput Library Preparation Kit, Millipore-Sigma, St. Louis MO); following this, enrichment and barcoding were done, and subsequently, libraries were sequenced using a 100bp paired-end run on an Illumina HiSeq2500 (Illumina, San Diego, CA).\n\nWhole-genome sequencing data for the 10 V. cholerae isolates were downloaded from NCBI’s SRA using the toolkit fastq-dump v2.9.3. An overview of the bioinformatics workflow adopted in this study has been provided (Figure 1).\n\nIS, insertion sequences; MLST, multilocus sequence typing; AMR, antimicrobial resistance; SNP, single nucleotide polymorphism; PATRIC, Pathosystems Resource Integration Center.\n\nUntrimmed sequence data quality reports were generated with FastQC v0.11.8 and MultiQC v1.7 using default settings.\n\nBacterial SNP calling was done using Snippy 3.2-dev, a tool for rapid and core genome alignments. V. cholerae genome assembly (accession number GCF_002892855.1) was obtained from NCBI’s nucleotide archive and used as a reference during variant calling. We then used BCFtools v1.9 to extract all SNPs that were shared by the 10 V. cholerae isolates. Custom bash scripts were used to extract only missense SNPs (nonsynonymous), available on GitHub (see Software availability)8. Gene ontology enrichment analysis was performed using PANTHER Overrepresentation Test (released 2019-06-06), annotation version PANTHER version 14.1 (Released 2019-03-12) and a reference list of V. cholerae. Biological process and molecular function enrichment analyses were also carried out using the same database.\n\nV. cholerae genomic reads were assembled using Unicycler v0.4.8-beta9 to generate contigs. The Pathosystems Resource Integration Center (PATRIC) v3.5.39 was used to annotate the assembled genomes.\n\nPATRIC v3.5.39 was used to generate genome assembly metrics, identify antimicrobial resistance genes, and virulence factors. We used ISfinder, a dedicated database for bacterial insertion sequences, to screen for the presence of insertion sequences in our assembled bacterial genomes. In addition, we performed a number of analyses using the different pipelines at the Center for Genomic Epidemiology to analyze the assembled bacterial genomes. These analyses were multilocus sequence typing (MLST) using MLST v2.0, plasmids searches using PlasmidFinder v2.0, phenotyping using BLAST-based on the V. cholerae database using MyDbFinder v1.2, and identification of acquired antibiotic resistance genes using ResFinder. For all the above pipelines, default settings were used.\n\nUsing the Mashtree command-line based tool, we generated a Newick file. This file was then uploaded to IcyTree, a browser-based phylogenetic tree viewer.\n\n\nResults\n\nThe 10 sequenced strains of V. cholerae belonging to Inaba and Ogawa serotypes were characterized through analysis of the whole-genome sequencing data. Except for one strain, which was a non-O1, all the other strains belonged to the serogroup O1 due to the presence of the rfbV-O1 gene. All 10 sequenced strains had biotype-specific genes ctxB, rstR, and tcpA; hence were all atypical EI Tor biotype variants of V. cholerae and these also belonged to the third wave of the seventh pandemic. In silico MLST revealed that the sequenced strains belonged to two different sequence types (STs); ST69 and ST515. Table 1 shows the genomic characteristics of the V. cholerae strains.\n\nIn addition, all the 10 sequenced strains were found to carry virulence-associated genes. These included MakA, ctxA, ctxB, carA, carB, trpB, clpB, ace, toxR, zot, rtxA, ompW, ompR, gmhA, fur, hlyA, and rstR. The 10 sequenced strains also carried genes for the Type VI secretion system (T6SS), including vasA, B, C, D, E, F, G, H, I, J, K, and L, vgrG-2, vgrG-3, vipA/mglA and vipB/mglB. The strains were also found to have the following genes: alsD (VC1589), involved in the synthesis of 2,3-butanediol; alsR, involved in the acetate-responsive LysR-type regulation; makA, the flagella-mediated cytotoxin gene; Type IV pilus genes, including tcpA, B, C, D, E, F, H, I, J, N, P, Q, R, S, and T, as well as icmF/vasK; adherence genes acfA, B, C, D, and IlpA; and quorum sensing system genes luxS, and cqsA. Pathogenicity islands were also present in all the 10 sequenced strains; namely, the Vibrio seventh pandemic islands VSP-1 and VSP-2 as well as VPI-1 and VPI-2.\n\nThe 10 sequenced strains all showed genotypic resistance to streptomycin, aminoglycosides, fosfomycin, fluoroquinolones, sulphonamides, trimethoprim, chloramphenicol/florfenicol, and tetracyclines, as illustrated in Table 2.\n\nFurthermore, all the 10 sequenced strains contained the VC1786 integrative and conjugative elements (VC1786ICE genes). Antimicrobial resistance genes associated with resistance to chloramphenicol, streptomycin, sulfamethoxazole, and trimethoprim usually found on the VC1786ICE, such as strA and strB, floR as well as sul2, were found present in the strains according to MyDbFinder 1.2. The sequenced strains were also found to have a genomic organization of the integrative and conjugative element similar to that of the V. cholerae ICEVchHai1 reference strain10.\n\nIn addition, all the sequenced strains had no plasmids according to PlasmidFinder 2.0, particularly the IncA/C plasmids, and cryptic plasmids pSDH1-2 were also absent in all the strains according to MyDbFinder 1.2 and in BLAST atlas.\n\nInsertion sequences were, however, present in all the sequenced strains; these included TS200/IS605, IS630, IS66, IS3, and IS4 (Table 3).\n\nKey: + / - Presence or absence\n\nSNP-based phylogenetic analysis showed an overall SNP difference of 120 among the 10 sequenced V. cholerae strains. Close relatedness was observed among strains SRR6871252, SRR6871253, and SRR6871254 (only seven SNP differences); strains SRR6871247, SRR6871249, and SRR6871250 (only four SNP differences), and among strains SRR6871245, SRR6871246, and SRR6871248 (only six SNP differences) (Figure 2).\n\nFurthermore, analysis for shared SNPs among the sequences showed that the sequenced strains shared 218 SNPs. Of these, 98 SNPs were missense (non-synonymous). Gene enrichment analysis of the SNPs using the PANTHER GO Ontology database showed enrichment in genes that mediate transmembrane-signaling receptor activity, peptidyl-prolyl cis-trans isomerase activity, and phosphor-relay response regulator activity.\n\n\nDiscussion\n\nThis was a cross-sectional study that aimed at providing a comprehensive genomic analysis of 10 whole-genome sequences of V. cholerae isolates collected during three different cholera outbreaks in Uganda between 2014 and 2016, submitted by the University of Maryland (Baltimore, MD, United States) to the NCBI SRA database under a study titled, “Molecular characterization of Vibrio cholerae responsible for cholera epidemics in Uganda by PCR, MLVA and WGS”.\n\nIn the genomic analysis, this study confirmed the identity of the isolates, provided the complete array of virulence genes, pathogenicity islands, antimicrobial resistance genes, integrative and conjugative elements, and antimicrobial resistance genes associated with these elements, plasmids, and insertion sequences. In addition, this study also provided a SNP-based phylogenetic analysis of the strains.\n\nThe identity of the isolates was in tandem with what was reported by Bwire et al., 2018. In addition, the finding of this study in regards to most of the isolates belonging to the O1 serotype are consistent with other studies elsewhere; these studies attributed this to the presence of the rfbV-O1 gene in isolates classified as O15. Unlike a similar study done in the East African region5 in which MLST revealed that their isolates belonged to a single ST (ST69), this study revealed that the isolates belonged to two STs; namely, ST69 and ST515. Strains of V. cholerae belonging to the ST515 have been reported elsewhere11.\n\nThe virulence genes reported in this study are similar to those reported in studies elsewhere5,12–14. These genes included, among others, those belonging to the Type IV secretion system, those involved in adherence, Type IV pilus genes, and those involved in quorum sensing.\n\nAccessory genetic elements, particularly pathogenicity islands previously reported to commonly occur in V. cholerae were also reported in this study; namely, VSP-1, VSP-2 as well as VPI-1 and VPI-215,16. These have not only been documented to encode virulence-associated genes in V. cholerae, but have also been reported to facilitate a better understanding of the evolutionary events that lead to the emergence of pathogenic V. cholerae clones15,16.\n\nDespite the World Health Organization (WHO) recommendations in regards to the management of cholera with oral rehydration salts in addition to antibiotics such as streptomycin, aminoglycosides, trimethoprim, fosfomycin, fluoroquinolones, sulphonamides, chloramphenicol/florfenicol, and tetracyclines, this study reports genotypic resistance in the isolates to these same antibiotics. Similar resistance has been reported in similar studies from the East African region and elsewhere5,17–19.\n\nThe presence of integrative and conjugative elements (VC1786ICE), containing resistance genes associated with sulfamethoxazole and trimethoprim, chloramphenicol, and streptomycin resistance, are also reported in this study. These are genomically similar to V. cholerae ICEVchHai117,19.\n\nThis study found no plasmids or intl genes. This could be attributed to the presence of integrative and conjugative elements, a factor that made them insignificant in regards to the encoding of antimicrobial resistance. Studies similar to this have reported similar findings5.\n\nThis study also reported the presence of insertion sequences IS605, IS66, IS3, and IS4. Insertion sequences have been described in various studies to be drivers of genetic variability. These studies have also alluded to them being fixed by natural selection each time that a mutation induced by these elements is selected20,21.\n\nThe presence of T6SS genes in the study isolates could explain their antimicrobial resistance gene profile. TS66-dependent killing of other bacteria is mostly directed to neighboring cells. These consequently release their DNA, which is ultimately taken up by the killer cells and in the process, these can integrate valuable genes including those that encode antimicrobial resistance. These may consequently evolve, leading to antimicrobial resistance in the killer cells22.\n\nThe results obtained from the SNP-based phylogenetic analysis show the relatedness of the Ugandan V. cholerae strains and are in agreement with the results obtained by Bwire et al., 2018.\n\nThe analysis for shared SNPs among the sequences and, consequently, gene enrichment revealed the enrichment in genes that mediate transmembrane-signaling receptor activity, peptidyl-prolyl cis-trans isomerase activity, and phosphor relay response regulator activity. These play a fundamental role in quorum sensing in V. cholerae, a process of cell-cell communication that allows these bacteria to share information about cell density and adjust gene expression accordingly23–25. Quorum sensing has been documented to regulate the expression of virulence factors in V. cholerae23–25.\n\n\nConclusions\n\nDespite the fact that bioinformatics capacity remains underdeveloped in Uganda and Africa as a whole, this study demonstrated the ability to apply bioinformatics approaches to zoom into genomes (in this case, V. cholerae genomes obtained from Uganda) to provide a comprehensive genomic analysis. This study sets a stage that encourages more sequencing work with potential public health consequences to be done in African settings. Furthermore, it also encourages the need to build bioinformatics capacity in African settings to enable analysis of whole-genome sequence data generated from the continent.\n\n\nData availability\n\nWhole-genome sequences of the ten Vibrio cholera isolates from Sequence Read Archive, Accession number SRP136117: https://identifiers.org/insdc.sra/SRP136117\n\nVibrio cholera reference genome assembly from NCBI Assembly, Accession number GCF_002892855.1: https://www.ncbi.nlm.nih.gov/assembly/GCF_002892855.1\n\n\nSoftware availability\n\n- Source code available from: https://github.com/gmboowa/extractonlymissenseSNPs-\n\n- Archived source code at the time of publication: https://doi.org/10.5281/zenodo.33544698\n\n- License: GPL-3.0", "appendix": "Grant information\n\nGerald Mboowa is supported through the DELTAS Africa Initiative [DEL15011] to THRiVE-2 (the Training Health Researchers into Vocational Excellence in East Africa). The DELTAS Africa Initiative is an independent funding scheme of the African Academy of Sciences’ (AAS) Alliance for Accelerating Excellence in Science in Africa (AESA) and supported by the New Partnership for Africa’s Development Planning and Coordinating Agency (NEPAD Agency) with funding from the Wellcome Trust [107742] and the UK Government.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nMengel MA, Delrieu I, Heyerdahl L, et al.: Cholera outbreaks in Africa. In: Cholera outbreaks. Springer; Curr Top Microbiol Immunol. 2014; 379: 117–44. PubMed Abstract \| Publisher Full Text\n\nBwire G, Orach CG, Abdallah D, et al.: Alkaline peptone water enrichment with a dipstick test to quickly detect and monitor cholera outbreaks. BMC Infect Dis. 2017; 17(1): 726. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKasolo F, Yoti Z, Bakyaita N, et al.: IDSR as a platform for implementing IHR in African countries. Biosecur Bioterror. 2013; 11(3): 163–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPallen MJ, Loman NJ, Penn CW: High-throughput sequencing and clinical microbiology: progress, opportunities and challenges. Curr Opin Microbiol. 2010; 13(5): 625–31. PubMed Abstract \| Publisher Full Text\n\nHounmanou YMG, Leekitcharoenphon P, Hendriksen RS, et al.: Environmental surveillance and genomic characterization of Vibrio cholerae O1 from fish, phytoplankton and water in Lake Victoria. In: 13th DWF Water Research Conference: Danish Water Forum. University of Copenhagen; 2019; 52. Reference Source\n\nBwire G, Sack DA, Almeida M, et al.: Molecular characterization of Vibrio cholerae responsible for cholera epidemics in Uganda by PCR, MLVA and WGS. PLoS Negl Trop Dis. 2018; 12(6): e0006492. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAruhomukama D, Sserwadda I, Mboowa G: Investigating colistin drug resistance: The role of high-throughput sequencing and bioinformatics [version 2; peer review: 2 approved, 1 approved with reservations]. F1000Res. 2019; 8: 150. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMboowa G: gmboowa/extractonlymissenseSNPs-: Extracts Genomic Variant Class and Counts (Version v1.0.0). Zenodo. 2019. http://www.doi.org/10.5281/zenodo.3354469\n\nWick RR, Judd LM, Gorrie CL, et al.: Unicycler: Resolving bacterial genome assemblies from short and long sequencing reads. PLoS Comput Biol. 2017; 13(6): e1005595. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nReimer AR, Van Domselaar G, Stroika S, et al.: Comparative genomics of Vibrio cholerae from Haiti, Asia, and Africa. Emerg Infect Dis. 2011; 17(11): 2113–21. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFu S, Hao J, Jin S, et al.: A Human Intestinal Infection Caused by a Novel Non-O1/O139 Vibrio cholerae Genotype and Its Dissemination Along the River. Front Public Heal. 2019; 7: 100. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCeccarelli D, Garriss G, Choi SY, et al.: Characterization of Two Cryptic Plasmids Isolated in Haiti from Clinical Vibrio cholerae Non-O1/Non-O139. Front Microbiol. 2017; 8: 2283. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAliabad NH, Bakhshi B, Pourshafie MR, et al.: Molecular diversity of CTX prophage in Vibrio cholerae. Lett Appl Microbiol. 2012; 55(1): 27–32. PubMed Abstract \| Publisher Full Text\n\nZhu J, Miller MB, Vance RE, et al.: Quorum-sensing regulators control virulence gene expression in Vibrio cholerae. Proc Natl Acad Sci U S A. 2002; 99(5): 3129–34. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFaruque SM, Mekalanos JJ: Pathogenicity islands and phages in Vibrio cholerae evolution. Trends Microbiol. 2003; 11(11): 505–10. PubMed Abstract \| Publisher Full Text\n\nMurphy RA, Boyd EF: Three pathogenicity islands of Vibrio cholerae can excise from the chromosome and form circular intermediates. J Bacteriol. 2008; 190(2): 636–47. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHendriksen RS, Price LB, Schupp JM, et al.: Population genetics of Vibrio cholerae from Nepal in 2010: evidence on the origin of the Haitian outbreak. mBio. 2011; 2(4): e00157–11. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHounmanou YM, Mdegela RH, Dougnon TV, et al.: Toxigenic Vibrio cholerae O1 in vegetables and fish raised in wastewater irrigated fields and stabilization ponds during a non-cholera outbreak period in Morogoro, Tanzania: an environmental health study. BMC Res Notes. 2016; 9(1): 466. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKaas RS, Ngandjio A, Nzouankeu A, et al.: The Lake Chad Basin, an Isolated and Persistent Reservoir of Vibrio cholerae O1: A Genomic Insight into the Outbreak in Cameroon, 2010. PLoS One. 2016; 11(5): e0155691. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSyvanen M: Insertion sequences and their evolutionary role. In: Bacterial Genomes. Springer; 1998; 213–20. Publisher Full Text\n\nSiguier P, Gourbeyre E, Chandler M: Bacterial insertion sequences: their genomic impact and diversity. FEMS Microbiol Rev. 2014; 38(5): 865–91. PubMed Abstract \| Publisher Full Text\n\nBorgeaud S, Metzger LC, Scrignari T, et al.: The type VI secretion system of Vibrio cholerae fosters horizontal gene transfer. Science. 2015; 347(6217): 63–7. PubMed Abstract \| Publisher Full Text\n\nRutherford ST, Bassler BL: Bacterial quorum sensing: its role in virulence and possibilities for its control. Cold Spring Harb Perspect Med. 2012; 2(11): pii: a012427. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nParker CT, Sperandio V: Cell-to-cell signalling during pathogenesis. Cell Microbiol. 2009; 11(3): 363–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWaters CM, Bassler BL: The Vibrio harveyi quorum-sensing system uses shared regulatory components to discriminate between multiple autoinducers. Genes Dev. 2006; 20(19): 2754–67. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "52002", "date": "19 Aug 2019", "name": "Jason Miller", "expertise": [ "Reviewer Expertise Genome assembly", "expression analysis." ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nSummary of the Manuscript: Genome assemblies of ten strains of Vibrio cholerae were generated from public sequence data. The assembled contigs were analyzed for gene content, insertion sequence content, SNPs content, etc.\n\nSummary of the Review: This reviewer is uncertain whether this manuscript contains publishable information. The experiment seems to be a repeat of a previously published experiment, using the same data exclusively. This experiment may have been designed as a test of the reproducibility of that previous study, but it is not presented as such. The Results section lists several ways that this study confirms findings in several previous publications but it leaves unclear what is novel. The Conclusions section of this study focuses on a different topic that was hardly addressed by the study: whether bioinformatics can be applied to genomic sequence, particularly in Uganda.\n\nMajor Revision:\n\nThe conclusions as stated are not supported by the data. Please refer to the conclusions section of the Abstract as well as the entire Conclusions section of the manuscript. These sections present a finding that researchers, particularly researchers in Uganda, can use bioinformatics to study cholera. If that truly was the research question, then the manuscript needs a new title and the report should be presented as a case study, N=1. In fact, the focus of the work seems to be cholera. If so, then the conclusions must pertain to cholera. If this was regionally the first study of its kind, then a sentence to that effect could be warranted in the Discussion section.\n\nThis manuscript needs to clarify whether this was a re-analysis of a previously published whole-genome sequencing project. The relationship to that previous study is easy to miss. Several sections are misleading in this regard. The Title refers to “Whole-genome sequence analysis” which usually includes sequencing. The Abstract refers to “The 10 sequenced strains” rather than “Ten previously-sequenced strains.” The Introduction says, “sequencing remains a less desirable option for most scientists in Uganda”, without pointing out that no sequencing was performed for this study. The Methods section says, “the whole-genome sequencing data was deposited in NCBI’s SRA with accession SRP136117” but that was the data of Bwire 2018; as a previously published fact, the deposition is not appropriate for this manuscript to report.\n\nThe manuscript must distinguish its novel findings from confirmations of previously published findings. This issue is addressed broadly in the Discussion section but not specifically in the Results. For example, the Results section says, “… all the 10 sequenced strains were found to carry virulence-associated genes. These included MakA, ctxA, ctxB, carA, carB, trpB, clpB, ace, toxR, zot, rtxA, ompW, ompR, gmhA, fur, hlyA, and rstR.” No prior work is cited there. The text fails to point out that the Bwire 2018 paper already reported that “All 63 isolates tested positive for ompW, toxR, and ctxA …” based on similar analysis of the very same sequencing reads.\n\nThe manuscript needs to explain why the re-analysis is worthy of publication. The Bwire 2018 paper already reported that the ten samples were Illumina-sequenced and assembled with the SPAdes software. The authors of that study posted ten assemblies at NCBI under BioProject PRJNA439310 and analyzed them in their publication. This manuscript presents ten assemblies of the same data using Unicycler (which uses SPAdes). Are the new assemblies different? Did the re-analysis highlight any flaws or shortcomings of the previous study? Did the re-analysis reveal anything new?\n\nMinor Revision:\n\nIn the Discussion, this study is self-described as a “cross-sectional study” which implies a look across samples from one timepoint. However, this study looked at three outbreaks over two years. Thus there was opportunity to include longitudinal and geo-spacial population analysis. This opportunity should be addressed in discussion, if possible.\n\nI presume the reads from the ten samples were assembled separately, but this is not stated in the Methods and not illustrated in Figure1 .\n\nThe phrase in the Abstract, “… in the strains”, should be “… in 10 strains isolated in Uganda” or similar. Basically, you cannot refer to “the” strains before saying which strains.\n\nWithin the Abstract, the background says one goal was to describe phylogeny by SNP-analysis but the abstract has no further mention of this. Either omit this from the background or include something about it in the results of the abstract.\n\nThe inline citations use notation like “Bwire 2018” but the references are numbered and listed in order of citation. This makes it hard to find the reference for a citation. Could the authors adopt one convention or the other?\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [] }, { "id": "53477", "date": "18 Sep 2019", "name": "Samit Watve", "expertise": [ "Reviewer Expertise Molecular genetics of Vibrio cholerae with a focus on Type six secretion", "natural transformation", "quorum sensing as well as expertise in genome sequencing and data analysis." ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nGeneral comments:\nThis study is tough to review because there is no clear scientific question being addressed and also lacks particularly novel or interesting findings. The authors claim, “The aim of this study was to provide the complete array of antimicrobial resistance genes, integrative and conjugative elements, virulence genes, pathogenicity islands, plasmids, and insertion sequences in the strains. In addition, this study also aimed to provide a single nucleotide polymorphism (SNP) based phylogenetic analysis of the strains.” What is unclear is what value this data array provides over the currently available genomic data. The study exclusively uses genomes that have been previously sequenced, assembled, analyzed and reported by Bwire et al., and are publicly available in the NCBI repository under accession number SRP136117. One could argue that having a pre-compiled dataset describing the presence/absence of antimicrobial resistance genes, integrative and conjugative elements, virulence genes, pathogenicity islands, plasmids, and insertion sequences enables faster access to important information for future researchers. If so, then why limit such reference dataset to a handful of genomes? A comprehensive database covering either all or a large proportion of the ~1300 or so publicly available V. cholerae genomes would be more appropriate for this kind of study. The original Bwire et al., study used genome sequencing as a tool to track disease spread across geography and time. This study uses the same information to report the presence of virulence, type VI secretion, type IV secretion and pathogenicity islands. These features are present almost universally in the seventh pandemic El Tor strains of Vibrio cholerae including the one in Uganda and this study does not add significantly to what’s already known about Vibrio cholerae biology. At best, it is a re-analysis of the Bwire et al. study with specific gene feature annotations extracted and tabulated. Since this study doesn’t represent an important advance to the state of knowledge in the field, in my view, this study does not merit publication.\n\nSpecific comments:\n\nThe section on sample collection and sequencing should be removed. Since the biological sample collection and sequencing was not performed in this study, it is inappropriate to include those methods here.\n\nBased on the workflow outlined in fig 1, it seems that for one aspect of the study (SNP variation analysis) the authors used the original RefSeq assemblies from Bwire et al., while for other aspects, such as looking for antibiotic resistance genes, finding plasmids etc. they generated new assemblies. The authors should explain why they used two different sets of assemblies and whether using different assemblies would generate significantly different results.\n\nFig 2 doesn’t have a scale bar for genetic distance or a distantly related outgroup to compare against. Therefore, the distance between two strains is impossible to determine.\n\nSome claims don’t make sense. For example, in the Discussion the authors claim, “In the genomic analysis, this study confirmed the identity of the isolates, …” or “The identity of the isolates was in tandem with what was reported by Bwire et al., 2018.” Since the genomes were from the strains that Bwire et al. sequenced, why was this ever in doubt?\n\nThe authors also claim, “The presence of T6SS genes in the study isolates could explain their antimicrobial resistance gene profile. TS66-dependent killing of other bacteria is mostly directed to neighbouring cells. These consequently release their DNA, which is ultimately taken up by the killer cells and in the process, these can integrate valuable genes including those that encode antimicrobial resistance. These may consequently evolve, leading to antimicrobial resistance in the killer cells.”\nSeveral studies have demonstrated that an active T6SS can lyse neighbouring cells, release their contents and allow for horizontal gene transfer. Even though this is a potential mechanism for horizontal transfer, there is no evidence that the presence of active T6SS in these strains has led to acquisition of antibiotic resistance genes. Indeed, Vibrio cholerae strains lacking an active T6SS can also have high efficiency of natural transformation and T6SS is by no means a requirement for horizontal gene transfer in Vibrio cholerae. Alternatively, antibiotic resistance genes can also be picked up through conjugation or transduction.\n\n“The analysis for shared SNPs among the sequences and, consequently, gene enrichment revealed the enrichment in genes that mediate transmembrane-signaling receptor activity, peptidyl-prolyl cis-trans isomerase activity, and phosphor relay response regulator activity. These play a fundamental role in quorum sensing in V. cholerae, a process of cell-cell communication that allows these bacteria to share information about cell density and adjust gene expression accordingly23–25. Quorum sensing has been documented to regulate the expression of virulence factors in V. cholerae23–25.”\nWhile it is true that the quorum sensing receptors LuxQ and CqsS have transmembrane-signaling and receptor activity as well as phosphorelay activity, the authors do not specify whether the luxQ and cqsS genes themselves carry any SNP’s. Many other two component signaling system proteins also have transmembrane-signaling and phosphorelay activity.\n\nPlease avoid vague or overly general sentences like: “Similar resistance has been reported in similar studies from the East African region and elsewhere5,17–19.” or “This study found no plasmids or intl genes… Studies similar to this have reported similar findings5.”\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1340
https://f1000research.com/articles/7-1755/v1	06 Nov 18	{ "type": "Research Article", "title": "Microbiota profiling with long amplicons using Nanopore sequencing: full-length 16S rRNA gene and whole rrn operon", "authors": [ "Anna Cuscó", "Carlotta Catozzi", "Joaquim Viñes", "Armand Sanchez", "Olga Francino", "Carlotta Catozzi", "Joaquim Viñes", "Armand Sanchez", "Olga Francino" ], "abstract": "Background: Profiling the microbiome of low-biomass samples is challenging for metagenomics since these samples often contain DNA from other sources, such as the host or the environment. The usual approach is sequencing specific hypervariable regions of the 16S rRNA gene, which fails to assign taxonomy to genus and species level. Here, we aim to assess long-amplicon PCR-based approaches for assigning taxonomy at the genus and species level. We use Nanopore sequencing with two different markers: full-length 16S rRNA (~1,500 bp) and the whole rrn operon (16S rRNA–ITS–23S rRNA; 4,500 bp). Methods: We sequenced a clinical isolate of Staphylococcus pseudintermedius, two mock communities (HM-783D, Bei Resources; D6306, ZymoBIOMICS™) and two pools of low-biomass samples (dog skin from either the chin or dorsal back), using the MinION™ sequencer 1D PCR barcoding kit. Sequences were pre-processed, and data were analyzed using the WIMP workflow on EPI2ME or Minimap2 software with rrn database. Results: The full-length 16S rRNA and the rrn operon were used to retrieve the microbiota composition at the genus and species level from the bacterial isolate, mock communities and complex skin samples. For the Staphylococcus pseudintermedius isolate, when using EPI2ME, the amplicons were assigned to the correct bacterial species in ~98% of the cases with the rrn operon marker, and in ~68% of the cases with the 16S rRNA gene. In both skin microbiota samples, we detected many species with an environmental origin. In chin, we found different Pseudomonas species in high abundance, whereas in dorsal skin there were more taxa with lower abundances. Conclusions: Both full-length 16S rRNA and the rrn operon retrieved the microbiota composition of simple and complex microbial communities, even from the low-biomass samples such as dog skin. For an increased resolution at the species level, using the rrn operon would be the best choice.", "keywords": [ "microbiome", "microbiota", "16S", "rrn operon", "nanopore", "canine", "low-biomass", "skin", "dog" ], "content": "Introduction\n\nThe microbiota profile of low-biomass samples such as skin is challenging for metagenomics. These samples are prone to containing DNA contamination from the host or exogenous sources, which can overcome the DNA of interest1,2. Thus, the usual approach is amplifying and sequencing certain genetic markers that are ubiquitously found within the studied kingdom rather than performing metagenomics. Ribosomal marker genes are a common choice: 16S rRNA and 23S rRNA genes to taxonomically classify bacteria3,4; and ITS1 and ITS2 regions for fungi5,6.\n\nUntil now, most studies of microbiota rely on second-generation sequencing (massive parallel sequencing), and target a short fragment of the 16S rRNA gene, which presents nine hypervariable regions (V1-V9) that are used to infer taxonomy7,8. The most common choices for host-associated microbiota are V4 or V1-V2 regions, which present different taxonomic coverage and resolution depending on the taxa9. V4 region represents better the whole bacterial diversity, although it fails to amplify Cutibacterium acnes (formerly known as Propionibacterium acnes), a ubiquitous skin commensal in humans. So, when performing a skin microbiota study, the preferred choice is V1-V2 regions, although they lack sensitivity for the Bifidobacterium genus and poorly amplify the Verrucomicrobia phylum10.\n\nApart from the biases derived from the primer choice, short fragment strategies usually fail to assign taxonomy reliably at the genus and species level. This taxonomic resolution is particularly useful when associating microbiota to clinics such as in characterizing disease status or when developing microbiota-based products, such as pre- or pro-biotics11. For example, in human atopic dermatitis (AD) the signature for AD-prone skin when compared to healthy skin was enriched for Streptococcus and Gemella, but depleted in Dermacoccus. Moreover, nine different bacterial species were identified to have significant AD-associated microbiome differences12. In canine atopic dermatitis, Staphylococcus pseudintermedius has been classically associated with the disease. Microbiota studies of canine atopic dermatitis presented an overrepresentation of Staphylococcus genus13,14, but the species was confirmed when complementing the studies using directed qPCRs for the species of interest13 or using a Staphylococcus-specific database and V1-V3 region amplification14.\n\nWith the launching of third-generation single-molecule technology sequencers, these short-length associated issues can be overcome by sequencing the full-length of the 16S rRNA gene (1,500 bp) or even the whole rrn operon (4,500 bp), which includes the 16S rRNA gene, ITS region, and 23S rRNA gene. The Oxford Nanopore Technologies MinIONTM sequencer is a single-molecule sequencer that is portable, affordable with a small budget and offers long-read output. Its main limitation is the high error rate.\n\nSeveral studies assessing the full-length 16S rRNA gene have already been performed using Nanopore sequencing to: i) characterize artificial and already characterized bacterial communities (mock community)15–17; ii) characterize complex microbiota samples, from the mouse gut18, wastewater19, microalgae20 and dog skin21; and iii) characterize the pathogenic agent in a clinical sample22–24. On the other hand, only two studies have been performed using the whole rrn operon to characterize mock communities25 and complex natural communities26.\n\nHere we aim to assess the potential of Nanopore sequencing using both the full-length 16S rRNA (1,500bp) and the whole rrn operon (4,500bp) in: i) a clinical isolate of Staphylococcus pseudintermedius, ii) two bacterial mock communities; and iii) two complex skin microbiota samples.\n\n\nMethods\n\nWe used two DNA mock communities as simple, well-defined microbiota samples:\n\n- HM-783D, kindly donated by BEI resources, containing genomic DNA from 20 bacterial strains with staggered ribosomal RNA operon counts (between 1,000 and 1,000,000 copies per organism per μl).\n\n- ZymoBIOMICS™ Microbial Community DNA standard that contained a mixture of genomic DNA extracted from pure cultures of eight bacterial strains.\n\nWe also sequenced a pure bacterial isolate of Staphylococcus pseudintermedius obtained from the ear of a dog affected with otitis.\n\nAs a complex microbial community, we used two DNA sample pools from the skin microbiota of healthy dogs targeting two different skin sites: i) dorsal back (DNA from two dorsal samples from Beagle dogs); and ii) chin (DNA from five chin samples from Golden Retriever/Labrador crossed dogs). Skin microbiota samples were collected using Sterile Catch-All™ Sample Collection Swabs (Epicentre Biotechnologies) soaked in sterile SCF-1 solution (50 mM Tris buffer (pH 8), 1 mM EDTA, and 0.5% Tween-20). DNA was extracted from the swabs using the PowerSoil™ DNA isolation kit (MO BIO) and blank samples were processed simultaneously (for further details on sample collection and DNA extraction see 27).\n\nThere were two ribosomal markers evaluated in this study: full-length 16S rRNA gene (~1,500 bp) and the whole rrn operon (~4,500 bp). Before sequencing, bacterial DNA was amplified using nested PCR, with a first PCR to add the specific primer sets (Table 1) tagged with the Oxford Nanopore universal tag and a second PCR to add the barcodes from the barcoding kit (EXP-PBC001). Each PCR reaction included a no-template control sample to assess possible reagent contamination.\n\nFor the first PCR, we targeted the full 16S rRNA gene using 16S-27F and 16S-1492R primer set and the whole rrn operon (16S rRNA gene–ITS–23S rRNA gene) using 16S-27F and 23S-2241R primer set (Table 1).\n\nPCR mixture for full-length 16S rRNA gene (25 μl total volume) contained 5 ng of DNA template, 5 μl of 5X Phusion® High Fidelity Buffer, 2.5 μl of dNTPs (2 mM), 1 μl of 16S-27F (0.4 μM), 2 μl of 16S-1492R (0.8 μM) and 0.25 μl of Phusion® Hot Start II Taq Polymerase (0.5 U) (Thermo Scientific, Vilnius, Lithuania). The PCR thermal profile consisted of an initial denaturation of 30 s at 98°C, followed by 25 cycles of 15 s at 98°C, 15 s at 51°C, 45 s at 72°C, and a final step of 7 min at 72°C.\n\nPCR mixture for the rrn whole operon (50 μl total volume) contained 5 ng of DNA template, 10 μl 5X Phusion® High Fidelity Buffer, 5 μl dNTPs (2 mM), 5 μl each primer (1 μM) and 0.5 μl Phusion® Hot Start II Taq Polymerase (1 U). The PCR thermal profile consisted of an initial denaturation of 30 s at 98°C, followed by 25 cycles of 7 s at 98°C, 30 s at 59°C, 150 s at 72°C, and a final step of 10 min at 72°C.\n\nThe amplicons were cleaned-up with the AMPure XP beads (Beckman Coulter) using a 0.5X and 0.45X ratio for the 16S rRNA gene and the whole rrn operon, respectively. Then they were quantified using Qubit™ fluorometer (Life Technologies, Carlsbad, CA) and volume was adjusted to begin the second round of PCR with 0.5 nM of the first PCR product or the whole volume when not reaching the required concentration (mostly for samples that amplified the rrn operon).\n\nPCR mixture for the barcoding PCR (100 μl total volume) contained 0.5 nM of first PCR product, 20 μl 5X Phusion® High Fidelity Buffer, 10 μl dNTPs (2 mM), and 1 μl Phusion® Hot Start II Taq Polymerase (2 U). Each PCR tube contained the DNA, the PCR mixture and 2 μl of the specific barcode. The PCR thermal profile consisted of an initial denaturation of 30 s at 98°C, followed by 15 cycles of 7 s at 98°C, 15 s at 62°C, 45 s (for the 16S rRNA gene) or 150 s (for rrn operon) at 72°C, and a final step of 10 min at 72°C.\n\nAgain, the amplicons were cleaned-up with the AMPure XP beads (Beckman Coulter) using a 0.5X and 0.45X ratio for the 16S rRNA gene and the whole rrn operon, respectively. For each sample, quality and quantity were assessed using Nanodrop and Qubit™ fluorometer (Life Technologies, Carlsbad, CA), respectively.\n\nIn most cases, the different barcoded samples were pooled in equimolar ratio to obtain a final pool (1000–1500 ng in 45 μl) to do the sequencing library.\n\nThe Ligation Sequencing Kit 1D (SQK-LSK108; Oxford Nanopore Technologies) was used to prepare the amplicon library to load into the MinIONTM (Oxford Nanopore Technologies), following the manufacturer’s protocol. Input DNA samples were composed of 1–1.5 μg of the barcoded DNA pool in a volume of 45 μl and 5 μl of DNA CS (DNA from lambda phage, used as a positive control in the sequencing). The DNA was processed for end repair and dA-tailing using the NEBNext End Repair/dA-tailing Module (New England Biolabs). A purification step using 1X Agencourt AMPure XP beads (Beckman Coulter) was performed.\n\nFor the adapter ligation step, a total of 0.2 pmol of the end-prepped DNA were added in a mix containing 50 μl of Blunt/TA ligase master mix (New England Biolabs) and 20 μl of adapter mix and then incubated at room temperature for 10 min. We performed a purification step using Adapter Bead Binding buffer (provided in the SQK-LSK108 kit) and 0.5X Agencourt AMPure XP beads (Beckman Coulter) to finally obtain the DNA library.\n\nWe prepared the pre-sequencing mix (14 μl of DNA library) to be loaded by mixing it with Library Loading beads (25.5 μl) and Running Buffer with fuel mix (35.5 μl). We used two SpotON Flow Cells Mk I (R9.4.1) (FLO-MIN106). After the quality control, we primed the flowcell with a mixture of Running Buffer with fuel mix (RBF from SQK-LSK108) and Nuclease-free water (575 μl + 625 μl). Immediately after priming, the nanopore sequencing library was loaded in a dropwise fashion using the SpotON port.\n\nOnce the library was loaded, we initiated a standard 48 h sequencing protocol using the MinKNOW™ software v1.15.\n\nThe samples were run using the MinKNOW software. After the run, fast5 files were base-called and de-multiplexed using Albacore v2.3.1. A second de-multiplexing round was performed with Porechop v0.2.330, where only the barcodes that agreed with Albacore were kept. Porechop was also used to trim the barcodes and the adapters from the sequences (Figure 1).\n\nMoreover, we removed 45 extra base pairs from each end that correspond to the length of the universal tags and custom primers. After the trimming, reads were selected by size: 1,200 bp to 1,800 bp for 16S rRNA gene; and 3,500 to 5,000 bp for the rrn operon. We mapped the sequences obtained to the rrn database using Minimap2 v2.931. Afterwards chimeras were detected and removed using yacrd v0.332.\n\nTo assign taxonomy to the trimmed and filtered reads we used to strategies: 1) a mapping-based strategy using Minimap231; or 2) a taxonomic classifier using What’s in my Pot (WIMP)33, a workflow from EPI2ME in the Oxford Nanopore Technologies cloud (based on Centrifuge software34).\n\nFor the mapping-based strategy, we performed Minimap2 again with the non-chimeric sequences. We applied extra filtering steps to retain the final results: we kept only those reads that aligned to the reference with a block larger than 1,000 bp (for 16S rRNA gene) and 3,000 bp (for the whole rrn operon). For reads that hit two or more references, only the alignments with the highest Smith-Waterman alignment score were kept. After filtering, the multimapping was mostly present in cases with entries that belonged to the same taxonomy.\n\nThe reference databases used in this study were:\n\n- Mock database: a collection of the complete genomes that were included in each mock community, as described by the manufacturer. The HM-783D database was retrieved from NCBI using the reference accession numbers, while Zymobiomics mock community has already its database online on the Amazon AWS server.\n\n- rrn database: sequences from the whole operon retrieved from Genbank25.\n\nFor the taxonomic classification using the WIMP workflow, which uses the NCBI database, only those hits with a classification score >300 were kept34.\n\nAn earlier version of this article can be found on bioRxiv (doi: https://doi.org/10.1101/450734)\n\n\nResults\n\nAfter Albacore basecalling and Porechop processing, we lost around 5% of the initial reads (3-13%). After length trimming step, we lost more sequences (Table 2). In general, the samples amplified using 16S rRNA marker gene recovered a higher percentage of reads after the quality control when compared to the rrn operon: 74–95% vs. 32–80%. Particularly for rrn operon, the largest percentage of reads was lost during the length trimming step: some of the reads included in that barcode presented the length of the 16S rRNA gene.\n\nZ1 and Z2 are replicates of ZymoBIOMICS™ Microbial Community DNA. HM, HM-783D mock community (BEI resources); Chin, microbiota from a pool of canine chin samples; Skin, microbiota from a pool of dorsal skin samples.\n\nAfter this first quality control, we performed an alignment with the mock and the rrn databases and checked for chimeras. Chimeras detected were dependent on the database used for the alignment. As a positive control, we used mock samples with their mock database. Chimera ratio was higher for 16S rRNA gene amplicons (around ~40%) than for rrn operon (~10%), suggesting that PCR conditions for the 16S rRNA gene need to be adjusted or the PCR cycles reduced.\n\nTo conclude, the final useful sequences when amplifying for either amplicon were ~40%. In 16S rRNA gene, sequences were lost in the chimera checking step. In the rrn operon, sequences were lost in the length trimming step, probably due to the underrepresentation of the amplicon in the flowcell, since we ran them together with full-length 16S rRNA amplicons in the same flow-cell.\n\nMicrobial Mock Community HM-783D contained genomic DNA from 20 bacterial strains with staggered ribosomal RNA operon counts (from 1,000 to 1,000,000 copies per organism per µl). The bacterial composition detected should be proportional to the operon counts. This mock community would allow us determining if our approach reliably represents the actual bacterial composition of the community, especially considering low-abundant species.\n\nWe analyzed HM-783D mock community against its own database, which contains only the 20 representative species. On the one hand, using 16S rRNA gene we were able to detect all the bacterial species present in the mock community, even the low-abundant ones. On the other hand, using the rrn operon we were able to detect only the most abundant species (at least 104 operon copies) (Figure 2). This could be due to the lower sequencing depth obtained with rrn when compared with 16S rRNA, and probably due to the underrepresentation of the rrn amplicon in the flowcell when running together with the full-length 16S rRNA amplicons in the same flow-cell, as detailed above. Moreover, the relative abundances of rrn operon sequences were more biased than those obtained from 16S rRNA gene sequencing, when compared to those expected, which confirmed that the primers for rrn need to be improved for universality.\n\nThe darkest blue represents the bacteria that were not detected (<104 copies with rrn operon), whereas the darkest red represents the most abundant bacteria.\n\nZymobiomics mock community presents the same amount of genomic DNA from 8 different bacterial species; the expected 16S rRNA gene content for each representative is also known, so we are able to determine if our approach represents the actual bacterial composition of the community reliably.\n\nBoth 16S rRNA gene and rrn operon sequencing were able to detect 8 out of 8 bacterial species for the Zymobiomics mock community, using Minimap2 and WIMP. The “Other taxa” group in Figure 3A can indicate: i) not expected taxa (wrongly-assigned species, or previous contamination); or ii) higher taxonomic rank taxa (sequences not assigned to species level).\n\n(A) Bar plots representing the relative abundance of the Zymobiomics mock community. “REF” bar represents the theoretical composition of the mock community regarding the 16S rRNA gene content of each bacterium. (B) Alpha diversity rarefaction plot using the rrn database at the species level. (C) WIMP output represented in a taxonomic tree and percentage of reads in each taxonomic rank.\n\nUsing the mock community database (that contains only the 8 members of that community), we aimed to assess the biases regarding the actual abundance profile. 16S rRNA gene better represented the bacterial composition of the mock community, when considering the abundances. The rrn operon amplification over-represented Escherichia coli and Staphylococcus aureus and under-represented Enterococcus faecalis.\n\nThe rrn database25 contains 22,351 different bacterial species, including representatives of the species in the mock community. When using the rrn database, we found that the rrn operon was a better marker than 16S rRNA: more than 98% of the sequences mapped to the corresponding species, and only <2% of the total sequences mapped to a wrong species with the rrn operon, whereas ~15% of the sequences were given the wrong taxonomy with 16S rRNA. We performed alpha diversity analyses using the same rrn database. The rrn operon hit 26 different species, whereas 16S rRNA over-estimated the actual diversity, with hits to 202 different species (Figure 3B). However, when considering abundances, the diversity values are more similar, with Shannon indices of 1.95 and 2.51, when using rrn operon and 16S rRNA, respectively (at 30,000 sequences/sample).\n\nUsing WIMP, we confirmed again the higher resolution power of rrn operon: ~70% of the sequences were assigned to the correct species compared to ~45% for the 16S rRNA gene. Among all the bacterial species included in the mock community, Bacillus subtilis presented the most trouble for the correct taxonomic classification. The theoretically expected abundance for B. subtilis is 17% using the 16S rRNA gene. When using WIMP, only 5% of the total sequences were correctly classified at the species level, another 5% was classified correctly at the genus level, and another 10% was incorrectly classified as other Bacillus species (Figure 3C).\n\nApart from the mock communities, we also sequenced an isolate of Staphylococcus pseudintermedius obtained from canine otitis. When using WIMP approach with rrn operon, 97.5% of the sequences were correctly assigned to the S. pseudintermedius. However, with the 16S rRNA gene, 68% of the sequences were correctly assigned at the species level and 13% at the genus (Table 3). The wrong assigned species for rrn operon was ~2.5%, compared to ~20% for the 16S rRNA gene. On the other hand, through mapping the sequences to rrn database using Minimap2, we obtained no hit to S. pseudintermedius, since there is no representative in the rrn database. Instead, they were hitting mostly to Staphylococcus schleiferi, which is a closely related species; there were also few hits to Staphylococcus hyicus and Staphylococcus agnetis. These results highlight the need of comprehensive databases that include representatives of all the microorganisms relevant to a microbiome to correctly assign taxonomy.\n\nAfter the first analyses with the mock communities, we were able to detect that the taxonomic resolution was higher when using rrn operon; however, the abundance profile was more reliable using 16S rRNA marker gene. If a bacterial species is not present in the database, the mapping strategy will give us the closest sequence resulting to an inaccurate taxonomic profile, such as we have seen for the Staphylococcus pseudintermedius isolate.\n\nHere, we aimed to taxonomically profile two complex and uncharacterized microbial communities from dog skin (chin and dorsal) using the two different markers and comparing the mapping strategy (Minimap2 and rrn database) with the WIMP workflow (NCBI database).\n\nFor chin samples of healthy dogs, we found a high abundance of Pseudomonas species followed by other genus with lower abundances such as Erwinia and Pantoea. Focusing on Pseudomonas, at the species level we were able to detect that the most abundant species was Pseudomonas koreensis, followed by Pseudomonas putida and Pseudomonas fluorescens (Figure 4A and Supplementary Table 1). On the other hand, dorsal skin samples were dominated by bacteria from the genera Stenotrophomonas, Sanguibacter, and Bacillus. We reached species level for Stenotrophomonas rhizophila and Sanguibacter keddieii. It should be noted that Glutamicibacter arilaitensis is the same species as Arthrobacter arilaitensis, but is the up-to-date nomenclature35 (Figure 4B and Supplementary Table 1). For both skin sample replicates, the results of the most abundant species converged and allowed for characterizing this complex low-biomass microbial community at the species level.\n\n(A) Chin samples: upper part of the graphic, bar plot of the composition at the genus level using WIMP (left) and Minimap2 (right); lower part, heat map of the Pseudomonas species within the community (scaled at 100%). (B) Dorsal skin samples: upper part of the graphic, bar plot of the composition at the genus level using WIMP (left) and Minimap2 (right); lower part, heat map of the ten most abundant taxa within the community. N/A, taxon was not present in the database.\n\nFinally, analyzing the dorsal skin samples, we also detected the presence of contamination from the previous nanopore run. We sequenced dorsal skin samples twice: one with a barcode previously used for sequencing the HM-783D mock community and another one with a new barcode (Table 2). We were able to detect mock community representatives within the re-used barcode (Figure 5). Some of them were found only in the sample that was using the re-used barcode (Sample_1); others were also present in the skin sample, such as Bacillus cereus or Staphylococcus aureus. In total, this contamination from the previous run was representing ~6% of the sample composition.\n\nSamples Skin_1 are the ones re-using the HM-783D barcode from the previous run within the same flowcell. Samples Skin_2 are using a new barcode (not used in a prior run of the same flowcell). Values within the heat map are relative abundances in percentages. The darkest blue represents the bacteria that were not detected and the darkest red the most abundant bacterial contaminants.\n\n\nDiscussion\n\nFull-length 16S rRNA and the rrn operon revealed the microbiota composition of the bacterial isolate, the mock communities and the complex skin samples, even at the genus and species level. Although Nanopore sequencing has a high error rate (average accuracy for the S. pseudintermedius isolate: 89%), we compensated this low accuracy with longer fragments to assess the taxonomy of several bacterial communities. In general, the longer the marker, the higher the taxonomical resolution both when using mapping software, such as Minimap2, or taxonomy classifiers such as WIMP in EPI2ME cloud.\n\nWhen using EPI2ME (WIMP with NCBI database), the amplicons from the S. pseudintermedius isolate were assigned to the correct bacterial species in ~98% and ~68% of the cases, using rrn operon and 16S rRNA operon, respectively. In a previous study, Moon and collaborators used the full-length 16S rRNA gene for characterizing an isolate of Campylobacter fetus and the marker assigned the species correctly for ~89% of the sequences using EPI2ME23. The ratio of success on the correct assignment at species level depends on the species itself and its degree of sequence similarity in the selected marker gene. Within the Staphylococcus genus, the 16S rRNA gene presents the highest similarity (around ~97%) when compared to other genetic markers36. On the other hand, we observed that using the mapping strategy (through Minimap2) could lead to a wrong assigned species if the interrogated bacterium has not any representative on the chosen database. This strategy provides faster results than EPI2ME, but it needs an accurate comprehensive and representative database.\n\nAnalyses of the mock communities allowed us to detect whether our approach represented the actual bacterial composition reliably. Moreover, with the HM-783D staggered mock community –with some low abundant species– we were able to detect the sensitivity of both approaches. When using the 16S rRNA marker gene, we were able to detect all bacterial members of both mock communities. However, when using the rrn operon, some of the low-abundant species were not detected. The likely reason is that we obtained a lower number of reads for this marker, up to one magnitude. Mock communities also allowed us to detect the potential biases of our primer sets for both markers, since some of the species detected were over- and under-represented. Actinomyces odontolyticus and Rhodobacter sphaeroides seem to not amplify properly, neither with 16S rRNA gene, nor the rrn operon. Previous studies also detected the same pattern for these specific bacteria even when using or comparing different primer sets16,21. Overall, the 16S rRNA primer set seemed less biased than rrn operon. When using the rrn operon, E. coli and S. aureus were overrepresented, whereas others were underrepresented, suggesting that the primers should be improved for universality.\n\nFocusing on the dog chin samples, we could detect that it was mostly Pseudomonas species that colonized: P. koreensis, P. putida, and P. fluorecens were the main representatives. Recently, Meason-Smith and collaborators found Pseudomonas species associated with malodor in bloodhound dogs37. However, these were not the main bacteria found within the skin site tested, but were in low abundance, differing from what we have found here. On the other hand, Riggio and collaborators detected Pseudomonas as one of the main genera in canine oral microbiota in the normal, gingivitis and periodontitis groups38. However, the Pseudomonas species were not the same ones that we have detected here. It should be noted that we had characterized these chin samples (and others) with 16S V1-V2 amplicons in a previous study27, where we found some mutual exclusion patterns for Pseudomonadaceae family. This taxon showed an apparent “invasive pattern”, which could be mainly explained for the recent contact of the dog with an environmental source that contained larger bacterial loads before sampling27. Thus, our main hypothesis is that the Pseudomonas species detected on dog chin came from the environment, since they have been previously isolated from environments such as soil or water sources39,40.\n\nThe most abundant species in dog dorsal skin samples were Stenotrophomonas rhizophila, Bacillus cereus, Sanguibacter keddieii, Sporosarcina psychrophila, Achromobacter xylosidans and Glutamicibacter arilaitensis. None of these specific bacterial species had previously been associated with healthy skin microbiota in human or dogs. Some of them have an environmental origin, such as Stenotrophomonas rhizophila, which is mainly associated with plants41; or Sporosarcina psychrophila, which is widely distributed in terrestrial and aquatic environments42. The Bacillus cereus main reservoir is also the soil, although it can be a commensal of root plants and guts of insects, and can also be a pathogen for insects and mammals43. Overall, environmental-associated bacteria have already been associated with dog skin microbiota and are to be expected, since dogs constantly interact with the environment27.\n\nRegarding Stenotrophomonas in human microbiota studies, Flores et al. found that this genus was enriched in atopic dermatitis patients that were responders to emollient treatment44. However, previous studies on this skin disease found Stenotrophomonas maltophila associated to the disease rather than Stenotrophomonas rhizophila45. Achromobacter xylosoxidans has been mainly associated with different kind of infections, as well as skin and soft tissue infections in humans46. However, both dogs included in this pool were healthy and with representatives of both genus/species, a fact that reinforces the need to study the healthy skin microbiome before associating some species at the taxonomic level to disease. The other abundant bacteria detected on dog skin have been isolated in very different scenarios: Sanguibacter keddieii from cow milk and blood47,48; and Glutamicibacter arilaitensis (formerly Arthrobacter arilaitensis) is commonly isolated in cheese surfaces35,49.\n\nFinally, some of the technical parameters used should be improved for better performance in future studies. In most cases we did not obtain enough DNA mass to begin with the indicated number of molecules for rrn operon amplicons. Thus, the flowcell contained an underrepresentation of rrn operon amplicons when compared to the full-length 16S rRNA gene. Moreover, in barcodes that contained rrn operon amplicons, a great percentage of reads were lost due to an inaccurate sequence size (~1,500 bp). One possible solution could be running each marker gene in different runs, so multiplexing samples with the same size amplicon to avoid underrepresentation of the larger one. When assessing chimera in mock samples using the specific mock database, we detected that the 16S rRNA gene formed a higher percentage of chimeras than rrn operon. Some options to improve that fact would include lowering PCR cycles performed. Better adjusting the laboratory practices would allow an increased DNA yield that meets the first quality control steps.\n\nTo conclude, both full-length 16S rRNA and the rrn operon retrieved the microbiota composition from simple and complex microbial communities, even from the low-biomass samples such as dog skin. Taxonomy assignment down to species level was obtained, although it was not always feasible due to: i) sequencing errors; ii) high similarity of the marker chosen within some genera; and iii) an incomplete database. For an increased resolution at the species level, the rrn operon would be the best choice. Further studies should be aiming to obtain reads with higher accuracy. Some options would include using the 1D2 kit of Oxford Nanopore Technologies, the new basecallers or the new flow cells with R10 pores. Finally, studies comparing marker-based strategies with metagenomics will determine the most accurate marker for microbiota studies in low-biomass samples.\n\n\nData availability\n\nThe datasets analyzed during the current study are available in the NCBI Sequence Read Archive, under the Bioproject accession number PRJNA495486: https://identifiers.org/bioproject/PRJNA495486.", "appendix": "Grant information\n\nThis work was supported by two grants awarded by Generalitat de Catalunya (Industrial Doctorate program, 2013 DI 011 and 2017 DI 037).\n\n\nSupplementary material\n\nSupplementary Table 1. Taxa found on skin microbiota of healthy dogs.\n\nList of all the taxa and their relative abundances found on dog skin microbiota samples (chin and dorsal skin). Results for both marker genes tested and for both approaches (Minimap2 + rrn database and WIMP with NCBI database).\n\nClick here to access the data\n\n\nReferences\n\nSalter SJ, Cox MJ, Turek EM, et al.: Reagent and laboratory contamination can critically impact sequence-based microbiome analyses. BMC Biol. 2014; 12(1): 87. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKong HH, Andersson B, Clavel T, et al.: Performing Skin Microbiome Research: A Method to the Madness. J Invest Dermatol. 2017; 137(3): 561–568. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLudwig W, Schleifer KH: Bacterial phylogeny based on 16S and 23S rRNA sequence analysis. FEMS Microbiol Rev. 1994; 15(2–3): 155–173. PubMed Abstract \| Publisher Full Text\n\nYarza P, Ludwig W, Euzéby J, et al.: Update of the All-Species Living Tree Project based on 16S and 23S rRNA sequence analyses. Syst Appl Microbiol. 2010; 33(6): 291–299. PubMed Abstract \| Publisher Full Text\n\nIwen PC, Hinrichs SH, Ruppy ME: Utilization of the internal transcribed spacer regions as molecular targets to detect and identify human fungal pathogens. Med Mycol. 2002; 40(1): 87–109. PubMed Abstract \| Publisher Full Text\n\nHibbett DS, Ohman A, Glotzer D, et al.: Progress in molecular and morphological taxon discovery in Fungi and options for formal classification of environmental sequences. Fungal Biol Rev. 2011; 25(1): 38–47. Publisher Full Text\n\nClarridge JE 3rd: Impact of 16S rRNA gene sequence analysis for identification of bacteria on clinical microbiology and infectious diseases. Clin Microbiol Rev. 2004; 17(4): 840–862. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJanda JM, Abbott SL: 16S rRNA gene sequencing for bacterial identification in the diagnostic laboratory: pluses, perils, and pitfalls. J Clin Microbiol. 2007; 45(9): 2761–2764. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWalters WA, Caporaso JG, Lauber CL, et al.: PrimerProspector: de novo design and taxonomic analysis of barcoded polymerase chain reaction primers. Bioinformatics. 2011; 27(8): 1159–1161. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKuczynski J, Lauber CL, Walters WA, et al.: Experimental and analytical tools for studying the human microbiome. Nat Rev Genet. 2012; 13(1): 47–58. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGrice EA: The skin microbiome: potential for novel diagnostic and therapeutic approaches to cutaneous disease. Semin Cutan Med Surg. 2014; 33(2): 98–103. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChng KR, Tay AS, Li C, et al.: Whole metagenome profiling reveals skin microbiome-dependent susceptibility to atopic dermatitis flare. Nat Microbiol. 2016; 1(9): 16106. PubMed Abstract \| Publisher Full Text\n\nPierezan F, Olivry T, Paps JS, et al.: The skin microbiome in allergen-induced canine atopic dermatitis. Vet dermatol. 2016; 27(5): 332–e82. PubMed Abstract \| Publisher Full Text\n\nBradley CW, Morris DO, Rankin SC, et al.: Longitudinal Evaluation of the Skin Microbiome and Association with Microenvironment and Treatment in Canine Atopic Dermatitis. J Invest Dermatol. 2016; 136(6): 1182–90. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLi C, Chng KR, Boey EJ, et al.: INC-Seq: accurate single molecule reads using nanopore sequencing. GigaScience. 2016; 5(1): 34. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBenítez-Páez A, Portune KJ, Sanz Y: Species-level resolution of 16S rRNA gene amplicons sequenced through the MinIONTM portable nanopore sequencer. GigaScience. 2016; 5: 4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrown BL, Watson M, Minot SS, et al.: MinIONTM nanopore sequencing of environmental metagenomes: a synthetic approach. GigaScience. 2017; 6(3): 1–10. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShin J, Lee S, Go MJ, et al.: Analysis of the mouse gut microbiome using full-length 16S rRNA amplicon sequencing. Sci Rep. 2016; 6: 29681. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMa X, Stachler E, Bibby K: Evaluation of Oxford Nanopore MinION Sequencing for 16S rRNA Microbiome Characterization. bioRxiv. 2017. Publisher Full Text\n\nShin H, Lee E, Shin J, et al.: Elucidation of the bacterial communities associated with the harmful microalgae Alexandrium tamarense and Cochlodinium polykrikoides using nanopore sequencing. Sci Rep. 2018; 8(1): 5323. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCusco A, Vines J, D’Andreano S, et al.: Using MinION to characterize dog skin microbiota through full-length 16S rRNA gene sequencing approach. bioRxiv. 2017. Publisher Full Text\n\nMitsuhashi S, Kryukov K, Nakagawa S, et al.: A portable system for rapid bacterial composition analysis using a nanopore-based sequencer and laptop computer. Sci Rep. 2017; 7(1): 5657. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMoon J, Kim N, Lee HS, et al.: Campylobacter fetus meningitis confirmed by a 16S rRNA gene analysis using the MinION nanopore sequencer, South Korea, 2016. Emerg Microbes Infect. 2017; 6(11): e94. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMoon J, Jang Y, Kim N, et al.: Diagnosis of Haemophilus influenzae Pneumonia by Nanopore 16S Amplicon Sequencing of Sputum. Emerg Infect Dis. 2018; 24(10): 1944–1946. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBenítez-Páez A, Sanz Y: Multi-locus and long amplicon sequencing approach to study microbial diversity at species level using the MinION™ portable nanopore sequencer. GigaScience. 2017; 6(7): 1–12. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKerkhof LJ, Dillon KP, Häggblom MM, et al.: Profiling bacterial communities by MinION sequencing of ribosomal operons. Microbiome. 2017; 5(1): 116. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCuscó A, Belanger JM, Gershony L, et al.: Individual signatures and environmental factors shape skin microbiota in healthy dogs. Microbiome. 2017; 5(1): 139. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZeng YH, Koblížek M, Li YX, et al.: Long PCR-RFLP of 16S-ITS-23S rRNA genes: a high-resolution molecular tool for bacterial genotyping. J Appl Microbiol. 2013; 114(2): 433–447. PubMed Abstract \| Publisher Full Text\n\nKlindworth A, Pruesse E, Schweer T, et al.: Evaluation of general 16S ribosomal RNA gene PCR primers for classical and next-generation sequencing-based diversity studies. Nucleic Acids Res. 2013; 41(1): e1. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWick R: Porechop. Reference Source\n\nLi H: Minimap2: pairwise alignment for nucleotide sequences. Bioinformatics. 2018; 34(18): 3094–3100. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMarijon P: yacrd: Yet Another Chimeric Read Detector for long reads. Reference Source\n\nJuul S, Izquierdo F, Hurst A, et al.: What's in my pot? Real-time species identification on the MinION. bioRxiv. 2015. Publisher Full Text\n\nKim D, Song L, Breitwieser FP, et al.: Centrifuge: rapid and sensitive classification of metagenomic sequences. Genome Res. 2016; 26(12): 1721–1729. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBusse HJ: Review of the taxonomy of the genus Arthrobacter, emendation of the genus Arthrobacter sensu lato, proposal to reclassify selected species of the genus Arthrobacter in the novel genera Glutamicibacter gen. nov., Paeniglutamicibacter gen. nov., Pseudoglutamicibacter gen. nov., Paenarthrobacter gen. nov. and Pseudarthrobacter gen. nov., and emended description of Arthrobacter roseus. Int J Syst Evol Microbiol. 2016; 66(1): 9–37. PubMed Abstract \| Publisher Full Text\n\nGhebremedhin B, Layer F, König W, et al.: Genetic classification and distinguishing of Staphylococcus species based on different partial gap, 16S rRNA, hsp60, rpoB, sodA, and tuf gene sequences. J Clin Microbiol. 2008; 46(3): 1019–1025. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMeason-Smith C, Older CE, Ocana R, et al.: Novel association of Psychrobacter and Pseudomonas with malodour in bloodhound dogs, and the effects of a topical product composed of essential oils and plant-derived essential fatty acids in a randomized, blinded, placebo-controlled study. Vet Dermatol. 2018. PubMed Abstract \| Publisher Full Text\n\nRiggio MP, Lennon A, Taylor DJ, et al.: Molecular identification of bacteria associated with canine periodontal disease. Vet Microbiol. 2011; 150(3–4): 394–400. PubMed Abstract \| Publisher Full Text\n\nPeix A, Ramírez-Bahena MH, Velázquez E: Historical evolution and current status of the taxonomy of genus Pseudomonas. Infect Genet Evol. 2009; 9(6): 1132–1147. PubMed Abstract \| Publisher Full Text\n\nMehri I, Turki Y, Chair M, et al.: Genetic and functional heterogeneities among fluorescent Pseudomonas isolated from environmental samples. J Gen Appl Microbiol. 2011; 57(2): 101–14. PubMed Abstract \| Publisher Full Text\n\nWolf A, Fritze A, Hagemann M, et al.: Stenotrophomonas rhizophila sp. nov., a novel plant-associated bacterium with antifungal properties. Int J Syst Evol Microbiol. 2002; 52(Pt 6): 1937–1944. PubMed Abstract \| Publisher Full Text\n\nYan W, Xiao X, Zhang Y: Complete genome sequence of the Sporosarcina psychrophila DSM 6497, a psychrophilic Bacillus strain that mediates the calcium carbonate precipitation. J Biotechnol. 2016; 226: 14–15. PubMed Abstract \| Publisher Full Text\n\nCeuppens S, Boon N, Uyttendaele M: Diversity of Bacillus cereus group strains is reflected in their broad range of pathogenicity and diverse ecological lifestyles. FEMS Microbiol Ecol. 2013; 84(3): 433–450. PubMed Abstract \| Publisher Full Text\n\nSeite S, Flores GE, Henley JB, et al.: Microbiome of affected and unaffected skin of patients with atopic dermatitis before and after emollient treatment. J Drugs Dermatol. 2014; 13(11): 1365–1372. PubMed Abstract\n\nDekio I, Sakamoto M, Hayashi H, et al.: Characterization of skin microbiota in patients with atopic dermatitis and in normal subjects using 16S rRNA gene-based comprehensive analysis. J Med Microbiol. 2007; 56(Pt 12): 1675–1683. PubMed Abstract \| Publisher Full Text\n\nTena D, Martínez NM, Losa C, et al.: Skin and soft tissue infection caused by Achromobacter xylosoxidans: report of 14 cases. Scand J Infect Dis. 2014; 46(2): 130–135. PubMed Abstract \| Publisher Full Text\n\nFernández-Garayzábal JF, Dominguez L, Pascual C, et al.: Phenotypic and phylogenetic characterization of some unknown coryneform bacteria isolated from bovine blood and milk: description of Sanguibacter gen.nov. Lett Appl Microbiol. 1995; 20(2): 69–75. PubMed Abstract \| Publisher Full Text\n\nIvanova N, Sikorski J, Sims D, et al.: Complete genome sequence of Sanguibacter keddieii type strain (ST-74). Stand Genomic Sci. 2009; 1(2): 110–118. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIrlinger F, Bimet F, Delettre J, et al.: Arthrobacter bergerei sp. nov. and Arthrobacter arilaitensis sp. nov., novel coryneform species isolated from the surfaces of cheeses. Int J Syst Evol Microbiol. 2005; 55(Pt 1): 457–462. PubMed Abstract \| Publisher Full Text" }	[ { "id": "40373", "date": "19 Nov 2018", "name": "Alfonso Benítez-Páez", "expertise": [ "Reviewer Expertise Human microbiome", "Microbial genomics", "Nanopore sequencing", "Computational biology", "Metagenomics" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nCusco and co-workers present an evaluation of both a mock community and the dog skin associated microbiota. The authors made use of the single-molecule Nanopore DNA sequencing technology and compared two different technical approaches by studying the nearly-full 16S rRNA bacterial gene and the nearly-full bacterial rRNA operon.\nIn my opinion, this work represents an important advance regarding the application of nanopore technology in the field of microbiome research.\nThe main strength of the work is its detailed technical description regarding the protocols for library preparation, sequencing and basecalling, that altogether facilitate the reproducibility. Moreover, the genetic data generated was properly deposited in a specialized database for public accession to whomever may want to replicate the analysis of long reads by similar approaches or new ones.\nThe figure quality is good and the information disclosed by them is well accompanied with appropriate captions.\nNotwithstanding, I have some minor concerns about the work that should be clarified, at least for me:\nThe last paragraph of page 7 describes the level of reads correctly assigned to species level for the microbial isolate Staphylococcus pseudintermedius. However, some of the values cited in the text do not match, at least, explicitly in Table 3. So, the authors should revise this issue or better describe the information obtained.\n\nThe authors found that the study of a nearly-full 16S rRNA gene reflects in a better way the expected abundances of microbial species present in the mock community tested. This comparative analysis with regard to the results obtained by using the rrn operon should be accompanied by a linear regression analysis, declaring respective Pearson's \"r\" coefficients, that can measure more accurately the efficiency of both methods and better support the authors' observations and conclusions.\n\nAdditionally to the observed richness (observed species) and Shannon diversity, the authors could also include a microbial community evenness evaluation of reference and observed microbiome data from the different approaches evaluated in the study, so that additional conclusions could be addressed.\n\nGiven the issues with underrepresentation of \"rrn\" data as a consequence of mixing this type of synthetic DNA with nearly-full 16S rRNA amplicons, the authors should highlight this observation as a major issue of this approach and state a clear recommendation to avoid this type of multiplexing for future studies.\n\nIt is necessary to better describe the contamination issues described in the last paragraph of the results (page 9). I'm not sure if this cross-contamination came from re-utilization of a flowcell or if this came from contamination of the barcoded-primer, used during nested PCR, with amplicons/DNA from the mock community. In a similar manner, the estimation of 6% of contamination has to be explained in detail (species/proportions discarded or having been taken into account to calculate this percentage).\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "43566", "date": "04 Feb 2019", "name": "Rasmus H. Kirkegaard", "expertise": [ "Reviewer Expertise microbial biotechnology", "nanopore sequencing" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nTitle: \"Microbiota profiling with long amplicons using Nanopore sequencing: full-length 16S rRNA gene and whole rrn operon\".\n\nSummary of the key results: The study demonstrates the use of nanopore sequencing for characterising low biomass samples with high levels of host DNA using a primer-based approach targeting the entire 16S rRNA gene or the 16S rRNA gene and the 23S rRNA gene.\nFurthermore, it evaluates the ability of these methods in the context of known references using mock communities and a pure culture using both the WIMP software and a custom mapping-based approach.\nThe study demonstrates that nanopore sequencing can give accurate classifications even at the current level of error rate if the reference database contains the right sequences. The study also shows how sequencing the longer fragment spanning both the 16S rRNA and 23S rRNA genes improves the taxonomic classification when the database contains a matching sequence.\nIs the work clearly and accurately presented and does it cite the current literature? The study mentions that the classification methods rely heavily on reference databases so it would be relevant to include citations for papers with methods for producing new reference sequences for both 16S rRNA and the longer fragment in the discussion (metagenomics, artificial long reads, primer free methods). Methods for improving read accuracy are also mentioned as important but the only methods mentioned are future upgrades from the company, relevant existing literature is not included (INC-seq, UMIs etc.). The study concludes that sequencing the entire “rrn operon” would be the best choice but it would be relevant to compare the size of current databases for the 16S rRNA gene versus the rrn operon. The presence of conserved sites for designing better primers is also extremely important but not discussed. Furthermore, there is evidence that quite a few organisms have unlinked rRNA genes, which will thus be missed by a full operon approach.\nCitations are also needed for bioinformatics tools for both processing and visualisation of the data.\nIs the study design appropriate and is the work technically sound? The study uses mapping to a reference database to point out that the sequences can get genus- and species-level classification. However, the method will always report a genus and a species even in the absence of the correct sequence in the reference database as indicated from the sequencing of the S. pseudintermedius pure culture with the “rrn” method. It will be important to simulate the impact on the results when there is no closely related sequences in the database. This could be done by removing all reference sequences within the Gammaproteobacteria and mapping the HM-783D to the modified database and monitor where the reads end up. It would also be helpful if there was a way to distinguish between reads that have the “correct” match and reads that just happen to map because the 16S rRNA gene is extremely conserved. Something similar would be relevant for the EPI2ME workflow but as the authors cannot control the reference database, it is probably not feasible. One of the advantages of the mock communities should be information about the copy numbers for the rRNA genes but there is no information on this included in the study and how it affects the results.\nAre sufficient details of methods and analysis provided to allow replication by others? The methods section lacks information about what happens after mapping the reads. How are the figures generated, what software is used, etc.? It would also be helpful if the specific scripts/commands used to run the bioinformatics analysis were available.\n\nFigures:\nFigure 1: bioinformatic workflow: The figure gives a decent overview of the bioinformatics processing but seems to miss the visualisation tools used. The main role of Albacore is basecalling the raw data not just demultiplexing. The figure could be improved further if you include the wet lab part of the work, so it becomes clear why the demultiplexing step is included and where the raw data comes from. A mapping step is integrated in the chimera detection (removal?) workflow but it might be better to omit mentioning mapping in that step as it can be confusing that the figure has two mapping steps in a row.\nFigure 2: heatmap mock community: The caption needs to explain what the numbers represent e.g. percentage of sequenced reads/mapped reads. It would be great if the heatmap included the “true” composition of the mock community for comparison. Copy number for each organism in the mock would also be relevant to include in the figure. Since there are only two columns, it would be better to have the sample labels at the top and with horizontal text preferably with a name that makes it easier to interpret the figure.\nFigure 3a: stacked bar chart: Even though stacked bar charts are very common it is not easy to read as they lack a common baseline for most of the values (See https://solomonmg.github.io/blog/2014/when-to-use-stacked-barcharts/ and https://peltiertech.com/stacked-bar-chart-alternatives/). I suggest that you use more of the heatmaps instead of introducing bar charts.\nFigure 3b: rarefaction curves: It would be great if you could add a dashed line for the expected “true” value for the mock community.\nFigure 3c: WIMP tree: This figure is quite complex to read. If the point with running both WIMP and a mapping-based approach with the two different amplicon types is to compare the methods, I suggest that you try to integrate the information better into one combined figure. This way you can help the reader to understand your message.\n\nFigure 4a: stacked bar chart+heatmap dog samples: Remove the stacked bar chart.\nFigure 4b: stacked bar chart+heatmap dog samples: Remove the stacked bar chart.\nGetting rid of the bar charts would allow for making a big heatmap with the data from Figure 4A and 4B combined. This way the reader can also compare the results from the two different sample sites. A naming system that makes it clearer that “_1” and “_2” are replicates would also help the reader interpret the figure. Presenting results aggregated at different levels, which could be included in one another is a bit confusing e.g., “Bacillus cereus” could be included in “Bacillus” which again could be included in “Bacillales”.\nFigure 5: heatmap mock community contamination: It is confusing that several cells in the heatmap have a value of “0” but with very different colours. Adding some meaningful labels with the contamination vs no contamination on the top could help the reader understand the figure without reading the caption.\n\nTables:\nTable 1: Primer sequences: Fine but could be moved to supplementary.\nTable 2: Samples and QC: Make headers easier to understand e.g. ”% seq 1st QC” could be “% of reads passing QC”, “Albacore pass” could just be “# reads after basecalling” etc. Where is the number after chimera detection?\nAdd a column with data accession ID and move the table to supplementary then the sample names can also be expanded so the reader does not have to look to the bottom for an explanation of abbreviations. I suggest adding a column at the end with the number of reads mapping/classified for each sample so the reader know what fraction is included.\nTable 3: Pure culture comparison WIMP vs. mapping: You need to make it clear in the table that Staphylococcus pseudintermedius is missing from the “rrn” database. As the paper mentions genus- and species-level classification as the target you may benefit from aggregating the values for S. pseudintermedius and S. pseudintermedius HKU10-03 as splitting this into strains makes it more confusing as your numbers in the text do not match the ones in the table.\nSupplementary Table 1: It would be great to include the mock communities in this table as well.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "43563", "date": "08 Feb 2019", "name": "Amanda Warr", "expertise": [ "Reviewer Expertise Genomics", "long read sequencing", "microbiome assembly" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nCusco et al. evaluate methods for long read sequencing and classification of marker genes from microbial communities, both for mock communities of known microbial composition and complex communities from dog skin from two anatomical locations, chin and back. They find that long read sequencing of 16S and the rrn operon is sufficiently accurate to classify microbes and that rrn is more sensitive at the species level.\nThis work demonstrates a valuable option for species identification from microbial samples using long reads, overcoming the current high error rate through covering a larger region. The work also highlights some of the issues that can arise from multiplexing amplicons of differing lengths when using Nanopore sequencing.\nOverall the paper is well written and detailed, however there are a few details I feel could be addressed:\n16S length reads in rrn barcodes: Do you expect this to be entirely from barcode misassignment or were these shorter fragments produced during PCR? You state that the loss of rrn amplicons during the length trimming step was probably due to over-representation of 16S amplicon on the flow cell, and most of the reads lost were roughly 16S amplicon sized - are you suggesting that there are large numbers of 16S reads that are assigned to rrn barcodes after 2 rounds of demultiplexing? Are these shorter reads actually whole 16S amplicons or fragments of rrn?\n\nExpected sensitivity given read count: The authors state that failure to detect the less abundant species from the mock community in the rrn dataset was \"probably\" due to their being fewer reads. As the proportions of the species in the mock samples are known, theoretically what total number of reads would be necessary to detect the less abundant species? Given the number of rrn reads obtained, did the authors detect as many species as they would expect to detect and what is the minimum total number of reads they would need to be likely to detect the lowest abundance species?\n\nDifferences in classification methods: Differences in classifications between the mock community database/rrn database and the NCBI database may be attributable to differences in the tools, with minimap2 being used for the mock and rrn databases and WIMP (based on centrifuge) being used for the NCBI database. My understanding is that the authors are mainly interested in classifications from different databases rather than differences between methods. While the authors do not directly compare the classification results between these different methods in text, some of the figures appear to imply that these results are directly comparable (e.g. Figure 3a). It would be useful if either all three databases were used with a single method (for example, using centrifuge with all three databases) or if these were at least more obviously separated or marked as coming from different classification methods in the figures.\n\nClassification rates against NCBI: The authors should further discuss ways to improve the classification rates, will the biggest improvements come from reduced error rate, better classification tools, improving species representation in databases? The authors conclude that in the future we should aim to improve accuracy, but one of the main results here is that sequencing the full 16S/rrn overcomes the problem of the current error rate - perhaps highlight benefits such as improved barcode assignment and emphasise that while this works well classification against a large database would likely improve with increased accuracy. The authors also conclude that rrn offers higher resolution at species level, however I suspect that currently more species have 16S sequences in databases than rrn.\nAdditionally, I have a few minor corrections mainly around small grammatical errors and figure/table modifications:\nPage 5: Paragraph beginning \"To assign taxonomy...\", change \"to strategies\" to \"two strategies\". Also I would change the last sentence on the page to say \"some of the reads excluded were the expected length of the 16S rRNA gene rather than the rrn operon\". Figure 1 should also be labelling Albacore as the basecaller.\n\nPage 6: change \"would allow us determining\" to \"would allow us to determine\".\n\nPage 11, column 2, line 2: change \"associated to\" to \"associated with\".\n\nFigure 3a would benefit from separating the reference bar from the other bars or adding this bar to the other two plots (currently it is grouped with Mock database, but it is also relevant to the rrn database and the NCBI database).\n\nFigure 4 text is quite difficult to read.\n\nTable 2: the title of the final column isn't clear. Is this the % of reads that pass the quality filters before chimera detection? Could another column be added showing number of reads that pass this filter?\n\nFigure 5: there are several different colours of 0 in this heat map?\n\nIn the conclusion the authors have suggested ways to improve accuracy of this method in the future, I would add the R2C2 method (Volden et al., 20181) as an option to improve consensus accuracy here also, while designed for cDNA it could be applied to fragments of genomic DNA.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "43564", "date": "18 Feb 2019", "name": "Kon Chu", "expertise": [ "Reviewer Expertise neuroinfection", "encephalitis" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe study compared the results of microbiota profiling using two different markers (16S rRNA and the rrn operon) and different classification methods. Because other reviewers have already made comprehensive reviews and comments including several critical points, I would like to add only a few minor points to the manuscript:\n\nFigure 2: according to the text, Actinomyces odontolyticus was detected using the 16S rRNA gene, however, '0' in the figure can create confusion. It would be better to represent the number of copies of Actinomyces odontolyticus using more decimal places or adding a caption for this species.\n\nFigure 3a:\n- It would be better to change the figure (e.g. heatmap) to make it easier for readers to recognize under-represented and over-represented bacteria. Listeria monocytogenes also seems under-represented in the analyses using the mock database and rrn database, and the corresponding sentence in Page 7 may be changed.\n- Include the classification method (WIMP, minimap2) along with the name of the database, as in figure 4, to allow general readers to more easily match the methods and the database.\n\nIn the last paragraph of page 7, it seems that the criteria of the percentage of wrongly assigned species for the rrn operon are different from that for the 16S rRNA gene.\n\nTable 3: I suggest making a caption for the difference between 'Staphylococcus' and 'Other Staphylococcus'.\n\nIf the authors would like to insist on better resolution by using the rrn operon, they need to demonstrate the data of the analysis using multiple species including species that tend to be under-represented or over-represented.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "43567", "date": "21 Feb 2019", "name": "Lee J. Kerkhof", "expertise": [ "Reviewer Expertise Molecular ecology of microbial systems" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nF1000 Research Cuscó et al. (https://doi.org/10.12688/f1000research.16817.1)\n\nComments to the authors:\nThe manuscript describes a study assaying 2 mock bacterial communities or 2 complex skin microbiome samples from dogs (chin or dorsal back) using both near-full length 16S rRNA genes and near-full length rRNA operons with the Oxford Nanopore MinION. The authors employ a library preparation method generating either 16S amplicons (1400 bp) or rRNA operons (4500 bp) including barcoding with a 1D ligation/sequencing kit and FLO-MIN 106 cells. The data analysis pipeline utilized Albacore basecalling, near-full length amplicon size selection, and screening by What’s in my Pot (WIMP) and Minimap2 against both NCBI and rrn databases. The authors demonstrate increased resolution at the species level with longer reads, that there can be large losses of raw sequence reads by size selection for rrn amplicons in their hands, and that the data analysis software and database can influence the results of MinION bacterial community analysis.\n\nIt would have been very helpful for the authors to put these findings into context with other papers in the literature using MinION and rRNA genes. For example, their results support what others directly sequencing near-full length16S amplicons (e.g. Shin et al. (20161), Mitsuhashi et al. (20172), and Benitez-Paez et al. (20163)) or rRNA operons (e.g. Benitez-Paez et al. (20174), Kerkhof et al. (20175)) have shown in mock communities or complex samples with respect to species-level resolution. Additionally, the screening of MinION reads with different 16S rRNA databases has also been described in the supplementary figures of Kerkhof et al. (20175). Likewise, an acknowledgement of the various software packages that has been employed to analyze the MinION reads in the scientific literature would benefit the readership. It appears that QIIME, BLASTN, Centrifuge, LAST aligner, Discontinuous MegaBLAST, WIMP, and MiniMap2 have all been used to identify OTUs for the MinION platform for 16S rRNA genes or rrn operons. As the authors have shown, the software/database being used can be very influential in the results of MinION screens and a synopsis of what they have found in context with other investigators (% bacterial assignment vs. % error) may point to a best practice for future studies.\n\nOther Specific Comments:\nPage 3: I find it awkward/confusing to indicate the number of operons per microorganism per microliter here for the mock communities. Bacteria generally have 1-15 ribosomal operons in their genomes. I think it is clearer to just indicate the number of target rRNA operons is 10^3-10^6 for this particular DNA mixture.\n\nPage 3: The barcoding expansion pack (EXP-PBC001) requires that the primers contain overhangs attached to the rRNA primers. This is not mentioned by the authors. Did they put overhangs on 27F/1492R/2241R? If so, the first round of target amplification may be affected by the presence of these overhangs. This should be indicated.\n\nPage 4: The authors clearly show the danger of performing PCR and only characterizing the amplification product by Qubit fluorescence. If they had done agarose gels on the PCR reactions, they may have detected the short amplification products in their initial rrn operon reactions. Furthermore, these short reads are preferentially ligated using the SQK-LSK108 sequencing kit since there are more picomole ends. This best practice of visualizing PCR amplifications for size determinations before sequencing should be explicitly stated.\n\nPage 4: I am a little confused by the 0.5 nM notation for PCR product in the barcoding reaction. If the authors used 50 microliter reactions, did they put 25 ng of 1st round PCR product in their barcoding reactions for a 15 cycle amplification? Can the authors just state the mass of DNA used to barcode? Secondly, Table 2 indicates BC1, BC2, and BC3 were not used. Was there a reason these barcodes were not utilized?\n\nPage 6: Stating that the rrn operon profiling was more biased probably because of the lower sequencing depth does not recognize that others have not reported comparable bias or that it is probably a reflection of their potentially compromised amplification efficiencies. This conclusion should be viewed with caution, given the amplification issues noted above.\n\nPage 11: The running of shorter (1500 bp) and longer (4500 bp) libraries on the same flow cell at the same time should enrich for the shorter reads. The MinION uses electrophoresis to move DNA molecules through the pores and smaller fragments should mobilize easier.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-1755
https://f1000research.com/articles/8-1295/v1	31 Jul 19	{ "type": "Software Tool Article", "title": "ChIPdig: a comprehensive user-friendly tool for mining multi-sample ChIP-seq data", "authors": [ "Ruben Esse" ], "abstract": "In recent years, epigenetic research has enjoyed explosive growth as high-throughput sequencing technologies become more accessible and affordable. However, this advancement has not been matched with similar progress in data analysis capabilities from the perspective of experimental biologists not versed in bioinformatic languages. For instance, chromatin immunoprecipitation followed by next-generation sequencing (ChIP-seq) is at present widely used to identify genomic loci of transcription factor binding and histone modifications. Basic ChIP-seq data analysis, including read mapping and peak calling, can be accomplished through several well-established tools, but more sophisticated analyzes aimed at comparing data derived from different conditions or experimental designs constitute a significant bottleneck. We reason that the implementation of a single comprehensive ChIP-seq analysis pipeline could be beneficial for many experimental (wet lab) researchers who would like to generate genomic data. Here we present ChIPdig, a stand-alone application with adjustable parameters designed to allow researchers to perform several analyzes, namely read mapping to a reference genome, peak calling, annotation of regions based on reference coordinates (e.g. transcription start and termination sites, exons, introns, and 5' and 3' untranslated regions), and generation of heatmaps and metaplots for visualizing coverage. Importantly, ChIPdig accepts multiple ChIP-seq datasets as input, allowing genome-wide differential enrichment analysis in regions of interest to be performed. ChIPdig is written in R and enables access to several existing and highly utilized packages through a simple user interface powered by the Shiny package. Here, we illustrate the utility and user-friendly features of ChIPdig by analyzing H3K36me3 and H3K4me3 ChIP-seq profiles generated by the modENCODE project as an example. ChIPdig offers a comprehensive and user-friendly pipeline for analysis of multiple sets of ChIP-seq data by both experimental and computational researchers. It is open source and available at https://github.com/rmesse/ChIPdig.", "keywords": [ "ChIP-seq", "read mapping", "peak calling", "genomic region annotation", "differential enrichment analysis", "heatmaps", "metaplots" ], "content": "Abbreviations\n\nbp: base pairs; ChIP-seq: chromatin immunoprecipitation followed by next-generation sequencing; cpm: counts per million; FDR: false discovery rate; GEO: Gene Expression Omnibus; GUI: graphical user interface; H3K36me3: histone H3 trimethylated at lysine 36; H3K4me3: histone H3 trimethylated at lysine 4; NCBI: National Center for Biotechnology Information; NGS: next-generation sequencing; PCR: polymerase chain reaction; PP: posterior probability; TMM: trimmed means of M values; TSS: transcription start site; UCSC: University of California Santa Cruz\n\n\nIntroduction\n\nInteractions between nuclear proteins and DNA are vital for cell and organism function. They control DNA replication and repair, safeguard genome stability, and regulate chromosome segregation and gene expression. Chromatin immunoprecipitation coupled with high-throughput sequencing (ChIP-seq) is a powerful method for assessing such interactions and has been widely used in recent years to map the location of post-translationally modified histones, transcription factors, chromatin modifiers and other non-histone DNA-associated proteins in a genome-wide manner. This progress has been fostered by the increasing technical feasibility and affordability of this technology, with more documentation and technical support available to researchers, including commercial library preparation kits, aided by the plummeting costs of sequencing and the ease of multiplexing samples. In light of this progress, ChIP-seq data sets are continuously deposited in publicly-accessible databases, such as the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) and the ENCODE consortium portal1. Therefore, there is an unprecedented wealth of epigenomic data in the public domain that can be used for integrative and correlative analyses.\n\nThis important advancement has not been accompanied with similar progress in user-friendly post-sequencing data analysis pipelines, which is still a significant bottleneck often handled by skilled bioinformaticians. A key step in ChIP-seq data analysis is to map reads to a reference genome assembly. Programs such as BWA2 or Bowtie/Bowtie23,4 are frequently employed for this purpose. Aligned data are then processed to find regions of enrichment along the genome, thereby identifying potential loci of DNA binding by the target protein or of deposition of the histone post-translational modifications of interest. This process is known as peak calling and can be performed by using algorithms such as MACS/MACS25,6 and SICER7. SAMtools is another popular software used in next-generation sequencing (NSG) analysis that provides various utilities for manipulating alignments, including sorting, merging, indexing and generating alignments in a per-position format8.\n\nFollowing the initial and common steps in ChIP-seq data analysis mentioned above, downstream analysis is often more customizable and may present a hurdle, especially in the case of comparing data derived from different conditions or experimental settings. Artifacts arising from bias of DNA fragmentation, variation of immunoprecipitation efficiency, as well as polymerase chain reaction (PCR) amplification and sequencing depth bias, result in ChIP-seq experiments with distinct signal-to-noise ratios and impose great challenges to the computational analysis9. Several methods addressing the differential enrichment analysis problem, i.e. the detection of genomic regions with changes in ChIP-seq profiles between two distinct samples or sets of replicate samples, have been proposed10–12. Generally, these methods rely on the initial detection of candidate peak regions by a conventional peak calling algorithm, and then this peak-defining information is applied to analysis with methods tailored for the differential expression analysis of RNA-seq data such as edgeR13 or DESeq14. Downstream analysis of ChIP-seq data may also involve the annotation of genomic regions based on reference coordinates (e.g. distance from nearest transcription start site) and visualization of normalized coverage by means of heatmaps and metaplots.\n\nImportantly, ChIP-seq data analysis typically relies, at least in part, on tools that have been designed to run primarily on Linux/Unix-based systems, while biologists who need to work with NGS may be unfamiliar with such operating systems. In the past few years, several packages designed to handle NGS data have been released for R, a popular platform-independent programming language and software environment for computing and graphics. Here we present ChIPdig15, an open-source application that leverages on several R packages to enable comprehensive and modular analysis of multi-sample ChIP-seq data sets through a user-friendly graphical user interface (GUI), allowing any experimental biologist with minimal computational expertise to use it easily.\n\n\nMethods\n\nChIPdig15 is developed in R using package Shiny and relies on multiple R/Bioconductor packages, including QuasR and BSgenome16, for mapping reads to a reference genome assembly17, BayesPeak for peak calling18, csaw12 and edgeR13 for differential enrichment analysis, ChIPseeker for annotation of genomic regions of interest19, and EnrichedHeatmap for generating coverage heatmaps20. Packages used in several analysis modules implemented in ChIPdig are GenomicRanges21, valr22, shinyFiles, GenomicFeatures, ggplot2, ggsignif, reshape2 and circlize. The GUI allows integrative and interactive usage of these powerful libraries without requiring programming or statistical experience from the user. The required packages may be installed manually by the user or automatically upon launching the tool for the first time. A step-by-step tutorial on how to use ChIPdig is provided at https://github.com/rmesse/ChIPdig.\n\nR and RStudio versions 3.4.1 and 1.0.44 or higher, respectively, are required, and both memory usage and execution times are contingent on the specificities of the intended analysis (e.g. size of input files, sequencing depths and size of the reference genome assembly).\n\nChIPdig has the following capabilities: (1) alignment of reads to a reference genome; (2) normalization and comparison of distinct ChIP-seq data sets; (3) annotation of genomic regions; (4) generation of comparative heatmaps and metaplots for visualization of normalized coverage in user-defined specific regions. Each capability corresponds to a specific analysis module which can be loaded separate from other modules and using different input files. For the alignment and annotation modules, the tool automatically fetches the desired reference assembly and gene model, respectively, from public repositories. Analysis of mammalian genome data is supported.\n\nAn earlier version of this article can be found on bioRxiv (DOI: https://doi.org/10.1101/220079).\n\n\nUse cases\n\nTo illustrate each module of the tool, raw single-end ChIP-seq data for H3K4me3 and H3K36me3 in the model organism Caenorhabditis elegans were downloaded from NCBI GEO with series IDs GSE28770 and GSE28776, respectively. These two histone modification marks have different distribution profiles, namely promoter-proximal enrichment with punctuated peaks for H3K4me3, and gene body enrichment with broad peaks for H3K36me323. The analysis was performed on a computer with Windows 7 Enterprise operating system (Microsoft Corporation, USA), 3.4 GHz CPU and 8 GB memory. Figure 1 shows the GUI displayed upon launching ChIPdig15 from R Studio. The left pane displays four radio buttons, each corresponding to one of the four analysis modules, as well a clickable box for selection of the folder containing the input files for the analysis.\n\nThe left pane displays four radio buttons for selection of the analysis module and a clickable box for selection of the folder containing the input files. Additional features are sequentially unlocked as the user progresses through the software suite. Outputs generated throughout the analysis and additional instructions are placed in the main panel under a tab corresponding to the analysis module selected by the user.\n\nThe aligning module of ChIPdig relies primarily on the QuasR package, which supports the analysis of single-read and paired-end deep sequencing experiments17. If desired, read preprocessing can be performed to prepare the input sequence files prior to alignment, e.g. removal of sequence segments corresponding to adapters and low-quality reads. Sequence files in FASTQ format for each input DNA and ChIP sample corresponding to the H3K4me3 and H3K36me3 ChIP-seq experiments were aligned to the WS220/ce10 reference genome (available through the BSgenome package16) (Figure 2), producing an average of 10 million mapped reads per sample. The execution time was approximately 2 h 30 min.\n\nThe user supplies a text file listing the input files corresponding to unmapped reads in either uncompressed (e.g.: '.fq', '.fastq') or compressed (e.g.: '.gz', '.bz2', '.xz') format. Both single-end and paired-end data are supported. A scroll-down menu displays the reference genome assemblies available for alignment.\n\nA recurrent problem in ChIP-seq data analysis lies at the comparison of multiple coverage profiles generated from different experiments or corresponding to different conditions. This problem is addressed in the second analysis module of ChIPdig. Upon its selection, the user is prompted to provide a tab-delimited text file with the names of each ChIP sample (treatment) mapped reads file in BAM format, the matching input DNA (control) file, a sample ID, the condition or target name, and a color designation for the output peak and coverage files (Figure 3a). Any of the 657 R built-in colors can be chosen. A bin size parameter set to 50 bp (base pairs) by default is used in both peak calling and genome-wide library normalization for differential enrichment analysis. If desired, duplicate reads can be removed prior to normalization and sequences can be extended to a median fragment length that can be estimated either computationally or experimentally (e.g. average size of fragments generated by chromatin shearing) (Figure 3b). The initial mapped read processing outputs a table with library sizes and median fragment sizes, a chromosome size list, and a multidimensional scaling plot representing the similarity of samples in the data set (Figure 3c). Data are normalized based on library sizes and on the trimmed means of M values (TMM) approach24, which is implemented in the edgeR package13. Initial mapped read processing for the H3K4me3 and H3K36me3 ChIP-seq data took approximately 10 min.\n\nThe user supplies a tab-delimited text file listing the input DNA and ChIP sample mapped read files (a). If desired, duplicate read removal and fragment size extension can be performed (b). The application outputs a table with library sizes and median fragment sizes, a chromosome size list, and a multidimensional scaling plot representing the similarity of samples in the data set (c).\n\nCompletion of the initial processing of mapped reads releases options for downstream processing which are posted to the sidebar panel, namely export of files representing sequencing coverage, peak calling and differential enrichment analysis (Figure 4). Coverage is expressed in log2-transformed counts per million (cpm) for each genomic bin and, for each sample ID indicated by the user, three files with bedGraph format are generated: treatment (ChIP sample), control (input DNA sample) and the treatment-to-control ratio. Each file can be loaded onto a genome browser for visualization (Figure 5). Generation of coverage files for H3K4me3 and H3K36me3 data took approximately 8 min.\n\nThe user may generate coverage files in bedGraph format, perform peak calling for each sample, and assess differential enrichment for a set of genomic regions of interest.\n\nValues represent log2-transformed counts per million (cpm) for the ChIP sample subtracted by log2-transformed cpm for the input sample. Files in bedGraph format were exported from ChIPdig and loaded onto the UCSC Genome Browser. These profiles exemplify the well-known scenario whereby H3K4me3 accumulates around the transcription start site and H3K36me3 is enriched at the gene body.\n\nPeak calling is performed via the BayesPeak package by using a hidden Markov model and Bayesian statistical methodology18. The user specifies a posterior probability (PP) threshold and genomic regions with PP above such threshold are identified as peaks. If replicates are indicated, commonly enriched genomic regions representing replicated peaks can be derived, and a track definition line can be added, allowing each file to be loaded onto the UCSC genome browser for visualization. In addition, a consensus peak set representing all candidate enriched regions across the full data set may be obtained. The user may be interested in choosing this option for posterior differential enrichment analysis. Peak calling was performed for the H3K4me3 and H3K36me3 ChIP-seq data sets, yielding 4994 replicated peaks for H3K4me3 and 25855 replicated peaks for H3K36me3. Peak calling execution time was approximately 12 h.\n\nDifferential enrichment analysis resorts to functions implemented in the csaw12 and edgeR13 packages. Following library size and TMM normalization, bin-level coverage is computed and, if desired, bins in which coverage for input DNA samples exceeds that of the corresponding ChIP samples are filtered off. If the user is interested in differential enrichment analysis in a specific set of regions, the corresponding file in BED format has to be provided. To illustrate this feature of ChIPdig, differential enrichment analysis was performed for comparing H3K4me3 coverage with that of H3K36me3 using either H3K4me3 replicated peak coordinates (Figure 6a) or those corresponding to H3K36me3 peaks (Figure 6b), with a false discovery rate (FDR) threshold of 0.1. As expected, at H3K4me3 genomic peak coordinates, H3K4me3 coverage is greater than that of H3K36me3, and the opposite is observed for H3K36me3 peaks. Differential enrichment analysis for both comparisons took approximately 5 min.\n\nH3K4me3 ChIP-seq coverage was compared with that of H3K36me3 using either (a) H3K4me3 peak genomic coordinates or those of (b) H3K36me3 peaks with a false discovery rate threshold of 0.1. For each analysis, a mean-different plot representing the library size-adjusted log2-transformed fold change (the difference) against the average log2-transformed coverage (the mean), as well as a box-and-whisker plot showing the global change in normalized coverage between the two conditions, were originated.\n\nAnnotation of genomic regions of interest is performed via the ChIPseeker package19. The user supplies the file with regions in BED format and chooses the reference genome assembly, as well as the distance upstream and downstream of the annotated transcription start site to be considered for assignment of promoter regions (Figure 7a). Replicated H3K4me3 and H3K36me3 peaks were annotated and, characteristically of these marks23, most H3K4me3 peaks were assigned to promoters (Figure 7b), whereas H3K36me3 peaks lie predominantly at gene bodies (Figure 7c). The execution time for both annotation operations was approximately 3 min.\n\n(a) The region file in BED format is uploaded and reference genome assembly, as well as distances upstream and downstream of transcription start site, are chosen. (b) Replicated H3K4me3 peaks are mostly promoter-proximal, whereas (c) H3K36me3 peaks are found predominantly at gene bodies.\n\nThe user can supply multiple coverage files in bedGraph format and generate heatmaps and metaplots to visualize coverage in a specific region set in a comparative manner. This module of ChIPdig relies on a custom algorithm which builds a coverage matrix based on bin size and reference coordinates (start, end or both) selected by the user. Such matrix is then plotted in the form of a comparative metaplot and a set of heatmaps. H3K4me3 and H3K36me3 coverage files were loaded onto ChIPdig, along with a BED file with C. elegans transcription units in the WS220/ce10 reference assembly. Coverage was computed in 25 bp bins and gene bodies were either expanded or compressed to 500 bp. Windows of 250 bp upstream and downstream of each region were selected (Figure 8a). Typical of H3K4me3, coverage is higher in the vicinity of the transcription start site, whereas H3K36me3 is enriched at transcription unit bodies (Figure 8b, c). Heatmap and metaplot generation took approximately 4 min.\n\nCoverage files for H3K4me3 and H3K36me3 were supplied to ChIPdig, as well as a file with coordinates of C. elegans transcription units. (a) The heatmaps and metaplot were generated by considering a 25-bp bin size, and gene bodies were resized to 500 bp. Windows of 250 bp upstream and downstream of each region were selected. As observed in both (b) the heatmaps and (c) the comparative metaplot, coverage for H3K4me3 is overall higher in the vicinity of transcription start sites, while that of H3K36me3 is enriched at transcription unit bodies.\n\nIn the past decade, several tools have been made available through Galaxy, an interactive system that provides a simple Web portal enabling users to analyze genomic data derived from high-throughput sequencing techniques25. We have compared results obtained by ChIPdig with those obtained within Galaxy using Bowtie3 for read alignment and MACS25,6 for peak calling. The percentage of mapped reads was higher for ChIPdig, but it should be noted that no read pre-processing was performed for read mapping in Galaxy (Figure 9a). For H3K4me3 peak calling, the number of called peaks using ChIPdig was similar to that in Galaxy, and, for H3K36me3, more peaks were called in ChIPdig than in Galaxy (Figure 9a). This can be attributed to the fact that the peak calling algorithm used by ChIPdig, and which is implemented in the BayesPeak package18, generates smaller peaks than those generated by Bowtie. More importantly, the spatial localization of ChIPdig-generated peaks overlapped very significantly with that of Galaxy-generated peaks (Figure 9b). We have also compared coverage tracks generated by ChIPdig (background-subtracted) with those generated by the Galaxy version of DeepTools and observed that the corresponding profiles are very similar (Figure 9c). These observations indicate that ChIP-seq data analysis using ChIPdig compares favorably with that in Galaxy.\n\n(a) The percentage of mapped reads using the alignment module of ChIPdig is higher than that using Bowtie within Galaxy, but it should be noted that read pre-processing was not included in the latter analysis. Peak calling in ChIPdig and in Galaxy resulted in similar numbers of peaks called for H3K4me3, and more H3K36me3 peaks were called using ChIPdig compared with Galaxy-generated peaks. (b) Importantly, ChIPdig-generated peaks overlapped very significantly with Galaxy-generated peaks. (c) Coverage tracks generated by ChIPdig have a profile similar to those generated by the Galaxy version of DeepTools.\n\n\nConclusions\n\nChIPdig15 is a user-friendly application for handling multiple ChIP-seq data sets and has diverse useful capabilities spanning a comprehensive analysis pipeline, namely: read alignment to a reference genome, ChIP-seq data normalization, peak calling, differential enrichment analysis, annotation of genomic regions, and generation of comparative heatmaps and metaplots for visualization of normalized coverage.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\n\nSoftware availability\n\nSource code available from: https://github.com/rmesse/ChIPdig.\n\nArchived source code at time of publication: https://doi.org/10.5281/zenodo.334578815.\n\nLicense: GNU General Public License 3.0.", "appendix": "Grant information\n\nThis research was supported by the GM107056 R01 grant awarded by the National Institutes of Health to Dr. Alla Grishok.\n\n\nAcknowledgements\n\nThe author would like to acknowledge Dr. Alla Grishok (Department of Biochemistry, Boston University School of Medicine, Boston, MA, USA) for appraising the work at its earliest stages.\n\n\nReferences\n\nDavis CA, Hitz BC, Sloan CA, et al.: The Encyclopedia of DNA elements (ENCODE): data portal update. Nucleic Acids Res. 2018; 46(D1): D794–D801. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLi H, Durbin R: Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics. 2009; 25(14): 1754–60. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLangmead B: Aligning short sequencing reads with Bowtie. Curr Protoc Bioinformatics. 2010; Chapter 11: Unit 11.7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLangmead B, Salzberg SL: Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012; 9(4): 357–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhang Y, Liu T, Meyer CA, et al.: Model-based analysis of ChIP-Seq (MACS). Genome Biol. 2008; 9(9): R137. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFeng J, Liu T, Qin B, et al.: Identifying ChIP-seq enrichment using MACS. Nat Protoc. 2012; 7(9): 1728–40. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZang C, Schones DE, Zeng C, et al.: A clustering approach for identification of enriched domains from histone modification ChIP-Seq data. Bioinformatics. 2009; 25(15): 1952–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLi H, Handsaker B, Wysoker A, et al.: The Sequence Alignment/Map format and SAMtools. Bioinformatics. 2009; 25(16): 2078–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMeyer CA, Liu XS: Identifying and mitigating bias in next-generation sequencing methods for chromatin biology. Nat Rev Genet. 2014; 15(11): 709–21. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiang K, Keles S: Detecting differential binding of transcription factors with ChIP-seq. Bioinformatics. 2012; 28(1): 121–2. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShao Z, Zhang Y, Yuan GC, et al.: MAnorm: a robust model for quantitative comparison of ChIP-Seq data sets. Genome Biol. 2012; 13(3): R16. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLun AT, Smyth GK: csaw: a Bioconductor package for differential binding analysis of ChIP-seq data using sliding windows. Nucleic Acids Res. 2016; 44(5): e45. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRobinson MD, McCarthy DJ, Smyth GK: edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010; 26(1): 139–40. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAnders S, Huber W: Differential expression of RNA-Seq data at the gene level – the DESeq package. EMBL, Heidelberg, Ger. 2012. Reference Source\n\nrmesse: rmesse/ChIPdig: First release of ChIP-dig (Version v1.0.0). Zenodo. 2019. http://www.doi.org/10.5281/zenodo.3345788\n\nBonhoure N, Bounova G, Bernasconi D, et al.: An introduction to the Biostrings / BSgenome framework. Nature. 2014; 11(1): 1–13.\n\nGaidatzis D, Lerch A, Hahne F, et al.: QuasR: Quantification and annotation of short reads in R. Bioinformatics. 2015; 31(7): 1130–2. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCairns J, Spyrou C, Stark R, et al.: BayesPeak--an R package for analysing ChIP-seq data. Bioinformatics. 2011; 27(5): 713–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYu G, Wang LG, He QY: ChIPseeker: an R/Bioconductor package for ChIP peak annotation, comparison and visualization. Bioinformatics. 2015; 31(14): 2382–3. PubMed Abstract \| Publisher Full Text\n\nGu Z: EnrichedHeatmap: Making Enriched Heatmaps.. R package version 1.6.0. 2017. Publisher Full Text\n\nLawrence M, Huber W, Pagès H, et al.: Software for computing and annotating genomic ranges. PLoS Comput Biol. 2013; 9(8): e1003118. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRiemondy KA, Sheridan RM, Gillen A, et al.: valr: Reproducible genome interval analysis in R [version 1; peer review: 2 approved]. F1000Res. 2017; 6: 1025. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiu T, Rechtsteiner A, Egelhofer TA, et al.: Broad chromosomal domains of histone modification patterns in C. elegans. Genome Res. 2011; 21(2): 227–36. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRobinson MD, Oshlack A: A scaling normalization method for differential expression analysis of RNA-seq data. Genome Biol. 2010; 11(3): R25. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGiardine B, Riemer C, Hardison RC, et al.: Galaxy: A platform for interactive large-scale genome analysis. Genome Res. 2005; 15(10): 1451–5. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "53175", "date": "07 Oct 2019", "name": "Tao Liu", "expertise": [ "Reviewer Expertise Bioinformatics algorithm development on NGS datasets" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAlthough the author explained the rationale of building the tool ChIPdig, I am not convinced why the author selected the exact tools to be included-- QuasR for alignment, BayesPeak for peak calling, differential calling with csaw+edgeR, ChIPseeker for genomic region annotation. To make a pipeline with pure R packages is an interesting idea. However, the end-users will miss the flexibility to choose tools they want -- usually the most popular tools in the field, compared with using the Galaxy or Snakemake framework.\nThe use case presented is only based on modENCODE C.elegans ChIP-seq datasets published in 2011, so I wonder how the pipeline can be scaled up to deal with bigger datasets or data from a larger genome such as human data (author can use recent ENCODE data as an example). Without more use cases on recent data, the tool seems yet another ChIP-seq pipeline tailored for reanalyzing specific datasets.\nLastly, ChIPdig uses QuasR qAlign function to align reads to reference genome, and qAlign is, in fact, a wrapper on Bowtie, which is included in Galaxy pipeline that the author compared. Therefore the difference of mapping results in figure 9a maybe just due to parameter settings. However, such information -- how the author generated Galaxy results -- is missing in the manuscript.\n\nIs the rationale for developing the new software tool clearly explained? Partly\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly", "responses": [] }, { "id": "55692", "date": "12 Nov 2019", "name": "Hinrich Gronemeyer", "expertise": [ "Reviewer Expertise Functional genomics and cancer" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nTechnological advancement over the past two decades has generated a scientific landscape that has shifted from gene-centric to genome-centric research. As consequence, we are facing an ever-increasing amount of (functional) genomics datasets in publicly available repositories, which is in principle an extremely rich asset of knowledge. Indeed, any research project should start with an analysis of the existing data and analyzed the obtained results in view of this - sometimes hidden - information. However, the main caveats are presently (i) the availability of simple user-friendly computational biology tools to query, analyze and integrate the existing data and (ii) the absence of quality assessment indicators attached to each data set.\n\nThe author of the present article has addressed the first issue for assays involving chromatin immunoprecipitation followed by sequencing of the co-IPed DNA (ChIP-seq). This is widely used for mapping the cistromes of transcription factors (TFs) and other chromatin binding factors and for monitoring DNA and functional histone modifications along the genome. He describes the \"ChIPdig\" tool, an R pipeline for ChIP-seq data analysis from the alignment of sequenced reads to the differential analysis of multiple datasets. The goal of this pipeline is fully justified, namely to provide users with a single pipeline that covers most of the required analytical steps without the need of jumping between tools available in different programming languages.\n\nUnfortunately, there are a number of issues which need to be addressed in order to make this tool attractive compared to others.\nConceptual issues. The author has assembled a pipeline which uses other R packages (described in 'Implementation') but does not consider their limitations.\nPeak calling. ChIP-seq patterns reflect the specifics of TF binding and DNA/histone modifications, which can be highly variable. in simple terms ChIP-seq signals acn be 'sharp' (e.g., TFs, H3K4me3) or broad (e.g., H3K27me3, H3K36me3). The various existing peak callers have very different characteristics and cannot be used without considering these aspects1.\n\nNormalization. ChIPdig uses the simplest way of normalizing data sets according to total mapped reads (TMRs). This is not state-of-the-art and can lead to problems when important differences exist between TMRs of different datasets. Thus, methodologies like quantile normalization perform significantly better2,3.\n\nComparison of H3K4me3 and H3K36me3 (Figs. 5, 6). While this is an illustrative example based on current knowledge in the field, the data comparison per se can be criticized, as very different types of signals (sharp H3K4me3 and broad H3K36me3 signals) are compared using the same peak caller and sub-optimal normalization. In Fig. 5 it appears that the control (input?) is simply subtracted (after normalization?) resulting in negative and positive (peaks/landscapes). It thus appears that the variation between the input and H3K36me3 signal intensity profiles is as high as the signals shown for H3K36me3. In Fig. 6 the value of these displays for functional interpretation is not clear.\n\nChIP-seq artifacts. The author addresses correctly the potential artifacts that can be obtained for a multiplicity of reasons in ChIP-seq experiments. Unfortunately, however, he does not offer solutions in the context of the ChIPdig pipeline. This is one of the most important issues concerning ChIP-seq data sets, as significant number of datasets cannot be used due to quality issues and very few tools incorporate quality assessment (e.g. qcGenomics; http://www.ngs-qc.org/qcgenomics/).\n\nChallenging examples. The examples that have been chosen to illustrate ChIPdig are the 'easy' ones, where good antibodies and thus, clean data sets exist. However, ChIPdig should be challenged with more 'difficult' TFs/histone marks and 'kinky' input profiles.\n\nTechnical issues. The main hindrance for non-bioinformatics trained scientists to exploit existing databases is the need for time-consuming data download and re-formatting. Currently only the qcGenomics database contains pre-computed data sets allowing for immediate access.\n\nComputational time. In addition to downloading datasets for analysis, users face as main technical disadvantages of ChIPdig the time for computation. As stated by the authors, alignment of 10 million reads take 2.5 hrs. This is for the C. elegans genome, so it would be far too long for the human genome.\n\nAlignment. It appears that the user has no choice on the alignment parameters and multiple alignment tools.\n\nOrganisms. A description is missing in the documentation whether ChIPdig is currently only available for C. elegans or whether it can be used for other reference genomes. In the present drop-down box, only C. elegans is available.\n\nMemory requirements. Trials with an 8Gb PC to use human ChIP-seq raw fastq files for trimming, etc. gave error messages and the program aborted.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Partly\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly", "responses": [] }, { "id": "59875", "date": "25 Feb 2020", "name": "Thomas Carroll", "expertise": [ "Reviewer Expertise Bioinformatics", "ChIP-seq", "ATAC-seq", "Splicing" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe author presents a R based toolset for the analysis of ChIP-seq data in a GUI framework. The construction of R based ChIP-seq analysis pipelines affords the potential for the use of wide range of tools from R and Bioconductor libraries while offering a low dependency piece of software.\nChIPdig uses QuasR, a wrapper for Bowtie, for the alignment of ChIP-seq data from a BSGenome object. The Rbowtie2 and Rsubread packages are now both available on Windows, Mac and Linux systems and should be considered alongside Bowtie. I believe they would offer significant speed and memory usage improvements over QuasR. Although these do not accept BSGenome objects, ChIPdig could easily generate the FASTA from these packages for use with indexing steps of both packages.\nBlacklisted regions should be considered in this tool as they have been shown to have strong effects on the QC, fragment length estimation and between sample normalisation. Inclusion of methods of blacklist filtering from known sources (such as Encode) or in software derived blacklists (using GreyListChIP) should be performed.\nThe output of BedGraph instead of BigWigs may cause some problems for users when working with larger genomes such as human or mouse. BigWigs may not be able to be exported on Windows systems but users of Mac or linux should have this option available to them to make this feature worthwhile.\nPeak calling is performed with BayesPeak. It is unclear how this performs on the different types of epigenetic marks used in this study. Some more options for peak calling could be included here to allow finer control of the stitching of peaks into larger peaks. A simple bin based peak calling approach such as implemented in the CSAW user guide would be useful here. How the identification of replicated peaks is not clear in text and could be expanded.\nThe example differential enrichment analysis compare H3k4me3 and H3K36me3 signals directly. This is a strange example as most differential ChIP-seq analysis is performed within the same antibody. An example comparing the change in one histone mark over different conditions/treatment/tissue types would be a more useful and relevant comparison.\nThis differential enrichment example does highlight a potential pitfall with this approach where the majority of sites change. The user should be warned in these circumstances as conclusions are likely to be invalid. Alternative normalisations such as to total mapped reads in peaks or total mapped reads to genome could be provided as options (as in Diffbind).\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Partly\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1295
https://f1000research.com/articles/8-1244/v1	30 Jul 19	{ "type": "Research Article", "title": "Levels of mercury in Nile tilapia (Oreochromis niloticus), water, and sediment in the Migori gold mining belt, Kenya, and the potential ramifications to human health", "authors": [ "Samwel Kola", "Laetitia Wakonyu Kanja", "James Mucunu Mbaria", "Joyce Gichiku Maina", "Mitchel Otieno Okumu", "Laetitia Wakonyu Kanja", "James Mucunu Mbaria", "Joyce Gichiku Maina", "Mitchel Otieno Okumu" ], "abstract": "Background: Understanding mercury levels in gold mining areas where locals consume fish is important in evaluating the risk to the population. This study determined the levels of total mercury (T-Hg) in Nile tilapia, water, and sediment in the Migori gold mining belt and the potential risk to human health.\n\nMethods: Water, sediment, and fish were sampled from 10 sites in Rongo and Nyatike and analyzed for T-Hg using cold vapor atomic absorption spectroscopy. Geo-accumulation index (IGeo) was used to evaluate sediment quality, and the estimated daily intake of fish per meal (EDIm), the target hazard quotient (THQ), and the maximum allowable fish consumption rate (CRmw) were calculated as health risk indices. Results: Sediment from 6 sites were moderately polluted with T-Hg, while 2 sites were strongly polluted. Water from all the sites had T-Hg levels higher than the FAO recommended level for surface water. About 78% (38/49) of all tissues sampled had T-Hg levels above the 0.5 µg/g limit for consumption by the general human population. About 31% (15/49) of muscle tissues had T-Hg levels above the 0.5 µg/g limit, while 88% (43/49) of brain, 69% (34/49) of liver, and 69% (34/49) of muscle tissues had T-Hg levels above the 0.2 µg/g limit for consumption by at-risk groups. There were positive correlations between T-Hg levels in brain, muscle and pond sediment, T-Hg levels in brain and water pH and temperature, and negative correlations between T-Hg levels in brain and levels in water, T-Hg levels in brain and weight of fish. EDIm and THQ ranged from 2.43-15.84 µg/g and 24.3-158.4 µg/g wet weight respectively while CRmw ranged from 1-4 fish.\nConclusions: Consumption of Nile tilapia from the studied area carries a significant risk of Hg exposure in frequent fish-eaters, pregnant women, and developing children, but is safe for the general population.", "keywords": [ "Gold mining", "Migori", "Nile tilapia", "Oreochromis niloticus", "total mercury" ], "content": "Introduction\n\nMercury (Hg) is naturally present in the environment (air, soil, water) in trace amounts1. The three predominant forms of Hg are; elemental mercury (Hg0), ionic/inorganic mercury (Hg (II)/ Hg2+) and organic mercury (alkyl mercury compounds; where methylmercury; MeHg is the most important)1. Hg is considered by the United Nations to be one of the worst environmental pollutants in the World2. It is distributed in the environment by both natural and human activities including mining and smelting3, volcanic eruptions4, combustion of fossil fuels5, solid waste compost6 and geological weathering7. The burden of Hg in the atmosphere is reported to be increasing at a rate of 1.5% annually8,9. Contaminated soil and water are the principal conduits through which environmental Hg gains access to the food chain, notably in plants and livestock, including fish8,9. Once in the food chain, Hg can bioaccumulate, leading to adverse outcomes on human health and even death9.\n\nAll of humanity is exposed to some level of mercury1. However, the occurrence and severity of adverse health effects following exposure to Hg depend on various factors including the chemical form, dose, route and duration of exposure as well as the age/ developmental stage of the person exposed1. Exposure via diet may also be significant, particularly when foods such as fish and seafood are contaminated with mercury1. Populations at risk of mercury contamination include pregnant women, new mothers, fetuses of women exposed to methylmercury1, individuals with diseases of the liver, kidney, nervous system, and lungs1, neonates who consume contaminated breast milk10,11, recreational anglers, subsistence fish farmers, and people who are frequent consumers of large amounts of fish or meat from marine mammals (seals and whales)1. Individuals with high occupational exposure, those who use consumer products laced with mercury (skin lightening creams and soaps), partakers of traditional ethnic medicines containing mercury, and individuals that use dental amalgams are all at an increased risk of exposure1.\n\nThe nervous system is the primary reservoir for environmental mercury or its compounds and produces symptoms such as depression, amnesia, insomnia, tremors, and hearing loss9. Notwithstanding, the systemic distribution of mercury has the potential to produce a range of symptoms in different body systems including the cardiovascular, renal, digestive, hematological, pulmonary, immune, and endocrine systems9. Significant effects on these organ systems include cardiomyopathy, tubular necrosis, inflammatory bowel disease, aplastic anemia, bronchitis, leukemia, and painful/irregular menstruation9. Others include miscarriage, spontaneous abortion, stillbirth, low birth weight, neural tube deficits, craniofacial malformation, delayed growth, and cerebral palsy9.\n\nSeveral countries and international organizations have established reference levels for daily or weekly methylmercury or mercury intakes based on risk assessments1. These levels provide estimates on the levels of mercury that are deemed to be safe. Reference intake levels for MeHg exposure range from 0.7-2 μg/Kg body weight per week1. Additionally, provisional tolerable weekly intakes (PTWIs) have been established by a joint FAO/WHO expert committee on food additives whose aim is to evaluate chemical contaminants in food1. This committee set the limit for total mercury at 5 μg/kg body weight and 1.6 μg/kg body weight for methylmercury1. According to the committee, the PTWI refers to the amount of a substance which can be consumed weekly over an entire lifetime without any significant risk to human health1. Moreover, the environmental protection agency of the United States has also developed a reference dose (RfDs) for mercuric chloride (0.3 μg/kg body weight/day) and methylmercury (0.1 μg/kg body weight/day)1. A reference concentration (RfC) for elemental mercury (0.3 μg/cubic meter) has also been developed by the same agency1. Exposures that are greater than the RfD imply that the potential for adverse health effects from mercury exposure also increases1. The Codex Alimentarius provides guideline levels of 0.5 mg for methylmercury per kg in non-predatory fish and 1 mg methylmercury/kg in predatory fish1. The United States Food and Drugs Agency has also set an action level of 1 mg methylmercury/kg in finfish and shellfish1. As for the European Union, a 0.5 mg mercury/kg has been set for fish, while Japan allows up to 0.4 mg total mercury/kg (or 0.3 mg methylmercury/kg) in fish1. Beyond these guidelines, some governments have also taken the initiative to provide dietary advice on the amount of fish to be consumed, the type of fish to be consumed and the groups of the human population that are at risk1.\n\nMigori County is home to a vibrant small-scale artisanal gold mining community where mercury (Hg) is used to extract gold (Au). Up to 40% of the Hg used is lost to the environment during the process12. Various studies from the area aimed at monitoring the health of the ecosystem have painted a picture of contaminated soil, streams, and fish2,12,13. The government of Kenya has recently launched a program dubbed the Economic Stimulus Program (ESP), which aims to boost fish production in the country by promoting inland fish farming14. The program is heavily dependent on fast-growing species of fish such as the Nile tilapia15 and has been rolled out in about 160/290 administrative units in Kenya including many within the Migori gold mining belt. Moreover, some of the fish farms have been set up near active gold mining areas and the levels of T-Hg in water from the springs, underground streams and Lake Victoria that supply these fish farms are relatively unknown. Nile tilapia is also a popular and highly consumed source of protein for the local population. These conditions in unison necessitate the present study, whose general objective was to investigate the levels of T-Hg in pond water, sediment, and tissues of farmed and wild-caught Nile tilapia in Migori County. Specific objectives were: i) to investigate the T-Hg contamination in tissues (brain, liver, muscle) of farmed Nile tilapia fish from Nyatike and Rongo sub-counties (Migori County) as well as captured Nile tilapia from Lake Victoria, Kenya, ii) to determine the T-Hg contamination in pond water, and sediment collected from the Migori gold mining belt and how this contamination relates to contamination in the fish and iii) to evaluate the risk to human health following dietary exposure to T-Hg in different tissues of Nile tilapia. The information generated from this study may be necessary for fish consumers and policymakers, both locally and internationally.\n\n\nMethods\n\nEthical approval was obtained from the Biosafety, Animal Care, and Use Committee of the Faculty of Veterinary Medicine, University of Nairobi (REF: FVM BAUEC/2018/148; Extended data, Figure S116). Guidelines on humane harvesting of fish17 were applied throughout the study. Extreme caution was exercised in handling concentrated mercury reagents and acids. Good laboratory practices (GLPs) were observed (use of gloves, gas masks, overalls, and fume chamber) at all times. Before the study began, it was ensured that the fume extraction system within the lab was in good working condition by following the laboratory’s standard operating procedures.\n\nPurposive sampling was used to collect fish, sediment, and water samples across sites within the Migori gold mining belt. The number of samples collected are summarized in Table S1 (Extended data16). The levels of total mercury in the collected samples was evaluated in the lab by following the methods of the United States Environmental Protection Agency (US-EPA) and Analytical Methods for Atomic Absorption Spectroscopy by Perkin Elmer18–20 with slight modifications. (Modifications: 1. the target weight for all samples was 0.5 grams but for those samples whose weights were below this weight, the whole sample was weighed and processed. 2. The reaction tube was elongated by fixing a glass (Soselex) column on top to prevent bumping and spill over.)\n\nThe Migori gold mining belt covers five sub-counties namely Suna West, Nyatike, Rongo, Kuria West, and Kuria East within Migori County (Figure 1). Apart from mining, other economic activities undertaken in the region include livestock, maize, tobacco, and sugar cane farming. The main rivers that drain the area are the Mara, Kuja, and Migori2. Rongo and Nyatike sub-counties were selected for sampling. This is because, in addition to being within the gold mining belt, inland fish farming is widely practiced in the region. Therefore, ten sites namely Minyenya, Kamagambo North, Masara, Nyabisawa, Ndiwa, Kokaka, Luanda Nyira, Kamagambo South, Siginga beach and Sori beach located within the two sub-counties (Figure 2), and bearing coordinates between 0°6’11.16’’-0°52’51.6’’S and 34°7’8.4-34°37’55.2’’E were selected as study sites.\n\nKey: 1: Nyatike, 2: Uriri, 3: Suna-East, 4: Suna-West, 5 Kuria-West, 6: Kuria-East, 7: Awendo, 8: Rongo.\n\nBy use of plastic trowels, near-surface (the top 5 cm) of fish pond sediments weighing approximately 1000 g was taken from each of the selected sites21, packed in plastic Biological Oxygen Demand (BOD) bottles (See details under extended data16. Table S2), labeled (location, sample type and the date of collection), kept on ice and transported to the laboratory where they were stored at -20° C until the time of analysis.\n\nTrace metal clean procedures, as described by Shafer, Shelton, and colleagues22,23 were used to collect water samples. Briefly, water samples were collected in 250 ml metal-free plastic bottles and acidified to a pH <2 using 69% ultra-pure nitric acid (HNO3) (Extended data16, Table S2) to prevent adsorption of potentially harmful elements onto the interior walls of the storage bottles as well and minimize microbial activity. The resultant mixture was then filtered through a 0.45 µm pore filter (Extended data16, Table S2) and stored in 125 ml metal-free plastic sample bottles and refrigerated at -20° C until the time of analysis. Mercury in the filtrate (0.45 μm), also referred to as “dissolved” mercury was of particular interest in this study as this is what may have had measurable biological effects on aquatic organisms23.\n\nFive tilapia fish (regardless of their sex) were captured from each site by using gill-nets (except Minyenya where four fish were sampled). Guidelines provided by the Faculty of Veterinary Medicine, University of Nairobi were adopted (REF: FVM BAUEC/2018/148; see Extended data16, Figure S1), and all efforts were taken to ameliorate harm to the animal model (fish). Additionally, guidelines on the humane harvesting of fish17 were adopted in euthanizing the collected fish.\n\nA two-step process involving electro-narcosis and asphyxiation was used17. Fish were initially stunned in an electric field of 2.5V/cm at 1000 Hz to make them insensible to pain17. The absence of eye-roll reflex when the fish were moved from side to side was used as a confirmation that insensibility had been achieved17. Death was then induced by asphyxiation in the air for 10 seconds and was confirmed by the lack of movement of the operculum17. Stunning fish by use of low current of electricity is considered a humane method of euthanasia as this current makes the fish insensible to pain. The length of the fish was measured by laying an extended tape measure on a flat surface and placing the fish next to the tape measure. The total length was determined by measuring from the tip of the fish’s mouth to the tip of the caudal fin. The weight of each fish was measured on an analytical balance (Sartorius: SECURA® 324-101N). A blade (number 11) was used to make a slit incision about 3–5mm in length through the ventral abdominal wall and a 10 cm2 sample of muscle tissue was taken from each fish21. Liver and brain tissues were also harvested by carefully dissecting them out from the fish and transferring the tissues to aluminum foil by use of gloves. The collected samples were then transferred to plastic sample bottles which were labeled, and packed in self-zipping polyethylene bags, stored in ice in an ice cool box and transferred to a -20° C freezer awaiting further analysis. The characteristics of the ecosystem in fish ponds in the study area are summarized under Table S3 (Extended data16).\n\nPre-treatment of equipment and sample bottles. Precautionary steps, as described by Shafer and others23 were taken before using the equipment or sample collection bottles. Briefly, all equipment used for sample collection and storage of sediment, water, and fish samples were pre-cleaned using high-purity nitric acid and rinsed with sufficient quantities of reagent water. This was done to ensure that they were free of trace metals. After cleaning, the bottles were stored in double-bagged zip-lock polyethylene bags to ensure that no detectable metal contaminants were present in the sampling equipment.\n\nReagents. All chemicals and reagents used were of analytical grade (Merck, Germany; Sigma-Aldrich, France; Central Drug House, India; Fisher Scientific, UK). (Table S2, Extended data16). Double distilled, de-ionized water was used for preparing working solutions and for all analytical work. Standard stock solutions of mercury were prepared from a high purity standard stock solution with a concentration of 1000 parts per billion (ppb) and were diluted to the corresponding mercury working standard solutions (i.e.10 ppb, 20 ppb, and 30 ppb). These working solutions were freshly prepared daily by diluting an appropriate aliquot of the stock solution using 1M hydrochloric acid (HCl; Sigma-Aldrich) and diluting the resulting solution to 100 ml with reagent water.\n\nStandard reference material for mercury in fish, i.e. the Community Bureau of Reference (BCR) – 463 (European Commission), was analyzed to ascertain the accuracy and precision of the experimental procedure. Alkaline solutions of sodium borohydride (NaBH4) were freshly prepared daily by dissolving 1.0 g of NaBH4 (Merck), and 0.25g of sodium hydroxide (NaOH) pellets (Merck) in 500 ml of distilled water. 3% v/v of HCl (Sigma-Aldrich) was used in the preparation of the carrier gas (Argon C-45). Stannous chloride (SnCl2) was freshly prepared by dissolving 62.5 g in 50 ml of 6 M HCl, the solution boiled for about 5 minutes, cooled, and nitrogen bubbled through it to expel any impurities of mercury. For sample digestion, 11 M nitric acid (HNO3; Merck), 18 M perchloric acid (HClO4; Merck) and HCl (Sigma-Aldrich) were used. Fused alumina anti-bumping granules (Merck) were used to avoid foam formation during sample digestion.\n\nEquipment and apparatus. All glassware used in the analysis were soaked overnight in 10 % (v/v) nitric acid (HNO3), followed by washing with 10% (v/v) hydrochloric acid (HCl). They were then rinsed with double-distilled water and dried before use. Samples were weighed on an analytical balance (Sartorius: SECURA® analytical balance: SECURA 324-101N), and sample digestion was carried out in a steam bath (DK Heating Digester from Velp Scientifica) in the confines of a fume hood. A Varian Model Spectr AA 220Z atomic absorption spectrometer equipped with a mercury hollow cathode lamp was used for the analysis of the total mercury content of samples. Flow injection and cold vapor generation were done via a Varian model vapor generation accessory (VGA) 77. The analytical wavelength and slit widths were 253.7 nm and 0.5 nm, respectively. The Varian model, Spectr AA 220Z software (Version: Spectr AA), was used to monitor the output.\n\nSample preparation and digestion. All samples, certified reference materials, standards, reagent blanks, and spiked samples were processed using slightly modified methods of the United States Environmental Protection Agency (US-EPA) and Analytical Methods for Atomic Absorption Spectroscopy by Perkin Elmer18–20. Briefly, a top pan analytical balance (Sartorius: SECURA®analytical balance: SECURA 324-101N) was calibrated each morning before weighing samples. Samples were removed from the freezer and allowed to thaw for about an hour. A batch of samples (approximately 20 in number) was digested simultaneously. Then, a 0.3–0.5 g aliquot of a well-homogenized sample was weighed and placed at the bottom of a glass digestion tube. For samples that were less than 0.5 g (particularly the brain and liver tissues), the weighing was done so as to obtain as much of the tissue samples as possible to enable processing. Nine milliliters of concentrated nitric acid (HNO3), 3 ml perchloric acid (HClO4) and 1 ml of HCl (to stabilize the pH of the matrix) were slowly added to the glass digestion tube in a fume hood. The tube was allowed to stand at room temperature (in the fume hood) until the initial reaction subsided (about 15 minutes). A spoon scoop of fused alumina anti-bumping granules (Merck) was added to the solution to prevent it from bumping and spilling over. A glass (Soselex) column was fixed on top of the glass digestion tube to prevent spilling over in the event of frothing. The tube was then placed on top of a steam bath unit (DK Heating Digester; Velp Scientifica) which was programmed to heat gradually to 150 °C over 10 minutes. This heating was maintained for 120 minutes to complete dissolution. The tube was then removed from the steam bath, and the solution allowed to cool to room temperature over 30 minutes. The solution was then carefully transferred into a 100 ml flat-bottomed volumetric flask, the tube and column rinsed thoroughly with small amounts of distilled water and the resultant contents transferred into the flat-bottomed volumetric flask. Six milliliters of saturated potassium persulfate (K2S2O8) and 30 ml of potassium permanganate (KMnO4) was added to the solution and slightly shaken to mix. The resultant was left to stand for 40 minutes. Additional portions of the KMnO4 solution were gradually added until the resulting purple color persisted for at least 15 minutes. After thorough mixing, 6 mL of sodium chloride-hydroxylamine sulfate was added to reduce the excess permanganate (this was confirmed by the color change from purple to colorless). Reagent water was then added to the mixture up to the 100 ml mark and treated samples were then filtered through grade 541 (diameter 110 µm) filter paper. Five milliliters of stannous sulfate was then added to each of the treated samples. Thereafter, each sample bottle was immediately attached to the aeration apparatus (one at a time) of the cold vapor atomic absorption spectrophotometer.\n\nAnalytical quality control. Precautionary steps were taken to rule out any interference that may have arisen in the course of running the analysis. Briefly, method blanks were analyzed to ensure that all the materials (solvents, reagents, glassware, and other sample processing hardware) were free from artifacts or interferences which may have had the potential to compromise the integrity of the analysis. Interference from sulfide was abolished by the use of 5% w/v potassium permanganate (KMnO4). Excess hydroxylamine sulfate reagent (about 25 mL) was used to ensure that free chlorine was absent before the mercury was reduced and swept into the cell. Also, the dead air space in the BOD bottle was purged before adding stannous sulfate.\n\nA preliminary run without reagents (using reagent water) was also used to rule out interference by volatile organic materials. Moreover, the accuracy of the procedure was determined by analyzing three certified reference materials (CRMs) namely; Tuna fish muscle fapas CRM, BCR 463 from Community Bureau of Reference, European Commission, vegetable puree CRM (EU) and fish CRM (EU). Recovery studies were performed by adding a known amount of standard solution of mercury chloride to some samples (spiked samples), which were then taken through the digestion procedure. The concentration of mercury in the resulting solutions was then analyzed.\n\nAnalysis of mercury in collected samples. The optimum operating temperature of the CVAAS instrument was set at 18 °C, and the circulating pump was adjusted to continuously pump at the rate of 1 L/min. Maximum absorbance was noted within 30 seconds. The bypass valve was then opened, and aeration continued until absorbance returned to the minimum value. The bypass valve was then closed, the fritted tubing removed from the BOD bottle, and aeration continued. The measurement time was set at 5 seconds, with a pre-reading delay of 45 seconds in between readings. Aliquots of 1.0, 2.0 and 3.0 mL of the mercury working standard (containing 0.1 mg/L or 1000 ppb) of mercury and 1 ml of fuming (37%) hydrochloric acid (HCl) solution was transferred to a series of 100 mL volumetric flasks and made up to the mark with reagent water. The standards had 10, 20, and 30 ppb of mercury, respectively. A calibration curve was automatically generated from the instrument’s software, plotting the absorbance of the standard against the concentration of mercury in parts per billion. The absorbance of the samples and standards were determined from the recording device and corresponding mercury concentrations tabulated.\n\nEvaluation of the degree of sediment contamination. The quantitative geochemical accumulation index (IGeo) was used to evaluate the level of mercury contamination in fish pond sediments collected from different sampling sites. This method applies the formula proposed by Müller24 to calculate the degree to which sediment is contaminated by mercury. Thus, IGeo= log2 (Cn/1.5×Bn), where; IGeo= the geochemical accumulation index, Cn=sediment metal concentration and Bn=geochemical background value of the metal. In this study, the global mercury background value was taken as 0.05, as described by Reimann25. Accordingly, mercury pollution in collected sediments was classified into seven categories (0–6) as: class 0 (unpolluted; IGeo≤0), class 1 (unpolluted to moderately polluted; 0≤IGeo≤1), class 2 (moderately polluted; 1≤IGeo≤2), class 3 (moderately to strongly polluted; 2≤IGeo≤3), class 4 (strongly polluted; 3≤IGeo≤4), class 5 (strongly to extremely polluted; 4≤IGeo≤5), and class 6 (extremely contaminated; IGeo>5)24.\n\nRisk-based consumption limits. Guidelines set by the United States Environmental Protection Agency (US-EPA)26,27 were used to calculate the potential health risk from consumption of Nile Tilapia sampled in the region. An assumption was made that the ingestion dose was equal to the absorbed dose of Hg, as has been described previously by Chien and colleagues28. Calculations on mercury consumption limits were based on the US-EPA reference dose (RfDo). The ratio between exposure and the reference dose indicated by the target hazard quotient (THQ), were calculated on the standard assumption of an integrated US-EPA risk analysis model. The methods described by Copat et al.29,30 were used to estimate the daily intake per meal (EDIm) and the target hazard quotient (THQ) as below;\n\n\n\n\n\nWhere EDIm is the estimated daily intake of mercury per meal size; MS is the standard weight portion of fish (230 g) for adults according to Hosseini et al.31; C refers to the concentration of mercury in mg/kg wet weight according to Marrugo-Negrete and colleagues32; BW is the body weight (assumed to be 70 kg for an adult)29. According to the US-EPA, the RfDo for T-Hg is 0.1 μg/g/day26.\n\nFor non-carcinogenic effects, the maximum allowable fish consumption rate in meals/week (CRmw) according to the US-EPA26 that would not be expected to cause any chronic systemic effects were calculated as below;\n\n\n\nWhere; MS is the standard weight portion of fish (230 g) for adults according to Hosseini et al.31, and C refers to the concentration of mercury in mg/kg wet weight according to Marrugo-Negrete and colleagues32. Considering an average adult body weight of 70 kg, the Hg USEPA Acceptable Daily Intake (ADI) can be approximated as 7 µg/day/adult (49 µg Hg/week)26,31.\n\nData for mercury analysis was expressed as the mean ± standard deviation. One-Way Analysis of Variance was used to analyze the levels of mercury in fish tissues across the sites. Tukey’s HSD test was used as a post-hoc test. Pearson’s correlation coefficient was used to determine whether there were any relationships between mercury levels in water and those in fish tissues, levels of mercury in sediment and those in fish tissues, and the levels of mercury in fish tissues and the pH of the pond water. The same test was also used to determine relationships between the levels of mercury in fish tissues and pond water pH, temperature, weight, and age of the fish. The t-test was used to analyze the relationship between the level of mercury and the type of fish culture practiced. Microsoft Excel (2016) and Statistical Package for the Social Sciences (SPSS, version 20.0) were used for statistical analysis. p≤0.05 was considered significant in all cases.\n\n\nResults\n\nThe mean levels of T-Hg were highest in the fish brain tissues and ranged from 0.128±0.021 µg/g wet weight (n= 5, 95% CI) at Nyabisawa in Nyatike sub-county to 3.128±1.421 µg/g wet weight (n= 4, 95% CI) at Minyenya in Rongo sub-county (Table 1 and Underlying data).\n\nValues are given as the mean of triplicate readings ± standard deviation (n=147). T-Hg: Total mercury\n\nThe fish muscle tissues were the second most contaminated with mean T-Hg levels ranging from 0.179±0.020 µg/g wet weight (n= 5, 95% CI) at Kamagambo South in Rongo sub-county to 0.917±0.099 µg/g wet weight (n= 5, 95% CI) in Minyenya. Table 1. The fish liver tissues were the least contaminated tissues with mean T-Hg levels ranging from 0.103±0.118 µg/g wet weight (n= 5, 95% CI) in Kamagambo South, Rongo sub-county to 0.588±0.374 µg/g wet weight (n= 5, 95% CI) in Kokaka, Rongo sub-county (Table 1).\n\nThere was a significant difference (p=0.00) in the mean levels of T-Hg in fish brain tissues sampled from the different sites (Table S4, Extended data16). Further evaluation of this variation revealed that the mean levels of T-Hg in fish brain tissues was significantly different between Nyabisawa and (Luanda Nyira, Kokaka, Ndiwa and Minyenya) (p=0.036), Kamagambo South and (Luanda Nyira, Kokaka, Ndiwa and Minyenya) (p=0.036), Sori beach and (Luanda Nyira, Kokaka, Ndiwa and Minyenya) (p=0.036), Siginga beach and (Luanda Nyira, Kokaka, Ndiwa and Minyenya) (p=0.036) and between Masara pond and (Luanda Nyira, Kokaka, Ndiwa and Minyenya) (p=0.036), Sori beach and (Luanda Nyira, Kokaka, Ndiwa and Minyenya) (p=0.036), and Kamagambo North and (Luanda Nyira, Kokaka, Ndiwa and Minyenya) (p=0.036) (Table S5, Extended data16).\n\nThere was a significant difference (p=0.00) in the mean levels of T-Hg in fish muscle tissues sampled from the different sites (Table S6, Extended data16). Further evaluation of this variation revealed that the mean levels of T-Hg in fish muscle tissues was significantly different between Kamagambo South and Minyenya (p=0.043), Kamagambo North and Minyenya (p=0.043), Kamagambo North and Masara pond (p=0.022), Kamagambo North and Siginga beach (p=0.022), Luanda Nyira and Masara pond, and Luanda Nyira and Siginga beach (p=0.022). (Table S7, Extended data16).\n\nThere was a significant difference (p=0.00) in the mean levels of T-Hg in fish liver tissues sampled from the different sites (Table S8, Extended data16). Post hoc analysis of this variation established that the mean levels of T-Hg in fish liver tissues was significantly different between Kamagambo South and (Sori beach, and Minyenya) (p=0.005), Masara pond and (Sori beach and Minyenya) (p=0.005), and between Ndiwa and (Sori beach and Minyenya) (p=0.005) (Table S9, Extended data16). The mean levels of T-Hg in all fish tissues (brain, liver, muscle) collected from the different study sites were found to be significantly higher than the critical value of 0.2 µg/g wet weight (n=49, 95%CI) which is the consumption limit for at-risk human populations (Table 3). However, only the fish brain tissues were found to have mean levels of T-Hg that were significantly greater than the critical value of 0.5 µg/g wet weight (n=49, 95%CI) which is the critical value for the general population (Table S10, Extended data16).\n\nThe levels of T-Hg in sediments ranged from 0.208±0.000 to 1.113±0.008 µg/g wet weight (n= 3, 95%CI) with six of the eight sample sites being moderately polluted (1≤IGeo≤2), whereas two sites (Minyenya and Kokaka; Rongo sub-county) were strongly polluted with total mercury (3≤IGeo≤4; Table 2).\n\nT-Hg: Total mercury, CI: Confidence interval, IGeo: geo-accumulation index.\n\nAn increase in the mean levels of T-Hg in the fish pond soil sediments coincided with an increase in the mean levels of T-Hg content in fish brain tissues (r= 0.528, sig. = 0.001, 95%CI) as well as an increase in the mean levels of T-Hg in fish muscle tissues (r= 0.524, sig. =0.001, 95%CI; Table S11, Extended data16). However, there was no significant correlation between the mean T-Hg levels in pond sediments and tilapia liver tissues (sig. =0.923, 95%CI; Table S11, Extended data16).\n\nThe mean levels of T-Hg in the water samples collected from the different sites ranged from 0.002±0.000 µg/ml to 0.004±0.001 µg/ml (n=3, 95% CI). (Table 3)\n\nThe water samples from all the sites were found to have mean T-Hg levels that were significantly greater than the critical value of 0.0001µg/ml (n=8, sig. = 0.000 at 95%CI) set by the food and agriculture organization for unpolluted surface waters33. (Table S12; Extended data16). An increase in the mean levels of T-Hg in fish pond water coincided with a decrease in the mean level of T-Hg in fish brain (r= - 0.402, sig. = 0.011, 95%CI) and muscle (r= - 0.616, sig. =0.000, 95%CI) tissues. (Table S13) Extended data. There was no significant correlation between the mean levels of T-Hg in pond water and the mean levels of T-Hg in fish liver tissues (sig. =0.874, 95%CI; Table S13, Extended data16).\n\nAn increase in the pH of fish pond water coincided with an increase in the mean levels of total mercury in fish brain tissues (r=0.48, p<0.05). (Table S14) Extended data. An increase in the temperature of the water in fish ponds coincided with an increase in the mean levels of total mercury in fish brain tissues (r=0.404, p<0.05; Table S14, Extended data16). The larger the fish, the lower the level of total mercury in fish brain tissues (r = -0.623; p<0.05; Table S14, Extended data16 and Underlying data34). There was no relationship between the age of fish and the level of total mercury in the fish brain, liver, and muscle tissues. (Table S14, Extended data16).\n\nThe type of fish culture practiced did not affect the mean levels of total mercury in fish brain tissues (Table S15, Extended data16). However, the type of culture practiced affected the mean levels of total mercury in fish muscle and liver tissues (Table S15, Extended data16).\n\nThe health risk indices from feeding on fish sampled from the different study sites are summarized in Table 3. Based on these results, feeding on fish sampled from Sori beach, Masara, Siginga beach, and Kokaka was associated with significant exposure to mercury as the CRmw was the least (Table 4). On the other hand, the risk of exposure to mercury was least in Kamagambo South on account of the large consumption rate in meals per week (Table 4).\n\nT-Hg: Total mercury, SD: Standard deviation, EDIm: Estimated daily intake of fish per meal, THQ: target hazard quotient, CRmw: Maximum allowable fish consumption rate in meals per week\n\n\nDiscussion\n\nFish can take up mercury from their environment and store it in relatively high concentrations in their tissues35. The levels of mercury recorded in fish tissues in this study were generally high when compared with the joint WHO/FAO critical reference guideline values for the general population36. Also, the levels of mercury, particularly in fish brain tissues were generally high when compared to the WHO/FAO critical reference values for at-risk populations (pregnant women, and children under five years)37. Our findings contradict those of Campbell and colleagues21 who reported that the levels of mercury in several species of fish captured from African Lakes, including Lake Victoria had mercury levels that were within WHO limits.\n\nIn this study, the levels of total mercury in fish tissues were in the order; brain>liver>muscle. This suggests that brain tissues of Nile tilapia in the region may have been releasing mercury at a slower rate than the liver and muscle tissues. It may also indicate that the distribution and accumulation of mercury in tissues of Nile tilapia seems to be biased towards the brain tissue. The role of metabolism on mercury bioaccumulation in the different tissues cannot be ruled out either. It could be that the metabolic rate in the brain tissues of Nile tilapia within the region was higher than other tissues and this may have predisposed fish brain tissues to a higher assimilation efficiency for methyl mercury than other tissues. Methyl mercury has a strong affinity for sulfhydryl groups in tissues and accumulates to a higher concentration in brain, muscle, and kidney tissues38.\n\nFrom the study’s findings, the fish brain may be the most dangerous part of the fish to eat as methyl mercury (a significant contributor of total mercury) is not eliminated from fish tissues by any practical cooking method38. This is worrying since the local community in Migori associate feeding on the fish brain with an increase in the intelligence quotient of the consumer.\n\nThe World Health Organization has developed guidelines on the maximum mercury intake per week, which serves as an advisory to fish consumers and in effect, help to avoid the accumulation of mercury in the body39. Our findings on risk-based consumption limits were specific for each tissue (brain, liver, and muscle). In reality, however, humans usually consume the whole fish. Therefore, considering the cumulative mercury concentration in the fish tissues (brain+liver+muscle), our results of the EDIm, THQ, and CRmw suggest that Kamagambo South and Nyabisawa were the areas where fish can be consumed with low risk to human health. Conversely, Masara, Kokaka, Siginga beach, and Sori beach are the areas where the risk to human health from consuming fish is highest.\n\nHigher levels of mercury in soil sediments coincided with higher levels of mercury in fish tissues (notably brain and muscle). According to Gupta and colleagues40, sediments are the most important reservoir of metals and other pollutants in the aquatic environment. The fact that heavy metal contamination in sediment has been shown to affect bioaccumulation of metals in aquatic organisms41 may partially explain why brain and muscle tissues in our study had high levels of mercury. However, it is unclear why there was no association between the levels of mercury in soil sediments and the levels of mercury in fish liver tissues. More studies may be needed to explore this phenomenon further.\n\nThe observation of low levels of total mercury in water sampled from fishponds may have something to do with the pH of the water, the presence of suspended solids in the ponds, as well as adsorption and precipitation processes. These factors have previously been shown to have the potential to remove metals such as mercury from solutions in the form of sulfides under anoxic conditions42. Also, inorganic mercury and methyl mercury (components of total mercury) have been reported to form complexes with naturally occurring dissolved organic carbon, thereby reducing the amount of mercury available in water43. This study revealed that as the content of mercury in water increased, the quantity of mercury in fish tissues (brain and muscle) decreased. According to past research, most of the mercury that accumulates in higher trophic level species originates from consumed food rather than direct aqueous accumulation44,45. However, this does not explain why there was a decrease in the quantity of total mercury in fish tissues (brain and muscle) as the level of mercury increased in the water. This observation warrants further research.\n\nThis study established that as the pH of pond water increased, so did the level of mercury in fish tissues (brain but not liver and muscle). Previous reports indicate that once mercury is assimilated by fish, it is distributed via the blood and stored in various tissues46. Thus, in the process of excretion, there is a transfer of mercury between ‘donor’ and ‘receiver’ organs, thereby implying that fish tissues are bound to have varying concentrations of mercury. Based on this, we speculate that in our case, alterations in the pH of fishpond water within the Migori gold mining belt appear to have favored bioaccumulation and distribution of mercury to the brain tissue as a receiver organ relative to liver and muscle tissues. Also, we posit that alterations in pH may also have influenced the release of mercury from brain tissues of Nile tilapia relative to liver and muscle tissues. The lack of a relationship between pH and mercury levels in liver and muscle tissues of Nile tilapia seems to suggest that the rate of bioaccumulation, distribution, and release from these tissues may not be dependent on changes in the pH of water in fishponds within the Migori gold mining belt. This study’s findings suggest that high water temperature was associated with higher levels of total mercury in fish brain tissues (but not liver and muscle tissues). Mechanistically, rising temperature in water may lead to an increase in the feeding rates of fish in response to higher metabolic demand47. Such an increase in food consumption could result in higher methyl mercury uptake and accumulation47. This may be due to a temperature dependent distribution and release of mercury from tissues of Nile tilapia in the study area. The findings suggest that the larger the fish, the lower the levels of mercury in fish brain tissues. Controversy abounds over the relationship between the weight of fish and the levels of mercury in fish tissues. Some studies have reported positive correlations between the two variables48 while others have reported negative correlations49. These differences may be due to differences in species, growth rates, body condition, and other bio energetic-related factors.\n\nThis study’s findings seem to suggest that the type of culture practiced in the Migori gold mining belt has some effect on the bioaccumulation and distribution of mercury in tissues of the Nile tilapia (particularly in liver and muscle tissues). The polyculture system (use of multi-species of fish) appears to be associated with high levels of mercury in fish tissues. It may be that polyculture practiced in fishponds within the Migori gold mining belt may expose Nile tilapia to unique physiological stressors that increase the rate of metabolism in fish tissues. Consequently, the rate of consumption of food among these fish rises resulting in a more pronounced uptake of mercury from food as well as a diminished elimination from the tissues. There is a need for studies to further investigate these findings.\n\nInterpretation of this study’s results may be limited by several factors. First, the statistical approach adopted may have ignored spatial correlations between sampling points and thus may have missed relevant information. Secondly, the fish diet has been reported by other studies to be one of the major pathways for the overall accumulation of mercury. In this study, the levels of mercury in feed and how potentially this would have translated to bioaccumulation in fish tissues were not investigated. Furthermore, this study used a single species of fish as a bio-indicator of pollution. Multi-species’ comparisons covering different feeding habitats of fish and a wide range of age categories may provide data that may facilitate stakeholders to distinguish recent exposure from the long-term load. Finally, this study did not explore the seasonal variation of mercury concentrations and what effect this might have had on bioaccumulation of mercury in the tilapia fish.\n\n\nConclusions\n\nLevels of mercury in tissues of Nile tilapia in the Migori gold mining belt are above-recommended limits. This study’s findings suggest that there is a significant health risk associated with mercury exposure through the consumption of Nile tilapia within the Migori gold mining belt. Thus, there is a need for fish consumption advisories about mercury in the Migori gold mining belt.\n\n\nData availability\n\nThis project contains the following underlying data:\n\nFigshare: Underlying data (1) of the study 'Levels of Mercury in Nile tilapia (Oreochromis niloticus), water, and sediment in the Migori gold mining belt, Migori, Kenya, and the potential ramifications to human health by Kola et al. https://doi.org/10.6084/m9.figshare.8869967.v350\n\nThis project contains the following underlying data:\n\nRaw data on the total mercury levels in water, sediment and tissues of Nile tilapia in the Migori gold mining belt.xlsx (Spreadsheet of recorded mercury levels)\n\nFigshare: Underlying data (2) on the study 'Levels of Mercury in Nile tilapia (Oreochromis niloticus), water, and sediment in the Migori gold mining belt, Kenya and the potential ramifications to human health by Kola et al. https://doi.org/10.6084/m9.figshare.8869964.v234\n\nThis project contains the following underlying data:\n\nRaw data on the weight and length of fish in the Migori gold mining belt.xlsx (Spreadsheet of recorded weights and lengths of captured fish)\n\nFigshare: Extended data on the study 'Levels of Mercury in Nile tilapia (Oreochromis niloticus), water, and sediment in the gold mining belt of Migori, Kenya, and the potential ramifications to human health' by Kola et al. https://doi.org/10.6084/m9.figshare.8870053.v416\n\nThis project contains the following extended data:\n\nKola-supplementary material.docx (word document containing supporting materials)\n\n○ The ethical approval form of the biosafety, animal use and ethics committee of the University of Nairobi. (Figure S1)\n\n○ The total number of sediment, water, and fish tissues collected from the ten sites in the study area. (Table S1).\n\n○ A summary of the reagents and chemicals used in the analysis. (Table S2).\n\n○ Characteristics of the ecosystems of the fish ponds where water, sediment, and fish were collected. (Table S3).\n\n○ Overview of the computed data (statistical analysis). Table S4 to Table S15.\n\nData are available under the terms of the Creative Commons Zero “No rights reserved” data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nThis study was funded by the Partnerships for Enhanced Engagement in Research (PEER) Science project and the United States Aid for International Development [PV 43358/34 awarded to SK]. The funds facilitated transport across the sites as well as collection and analysis of data.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nThe authors would like to acknowledge the staff of the Migori County Fisheries Department notably Ms. Ruth, Mr. Joseph Oruru and Mr. George for their assistance in the identification of sites, and facilitating coordination with the fish farmers as well as sample collection. Special gratitude is extended to technical staff at the Department of Public Health, Pharmacology and Toxicology, University of Nairobi notably Mr. Joseph Nderitu, Mr. Nduhiu Gitahi, Mr. Maloba and Mr. Kimotho for facilitating the storage and processing of the collected fish samples. Also, we appreciate the invaluable help and guidance offered by staff at the analytical chemistry laboratory of the Kenya Plant Health Inspectorate (KEPHIS) who contributed to the successful analysis of the collected fish samples. Lastly, we would like to thank Dr. Gerald Muchemi, Dr. Florence Mutua and Mr. Billy Odhiambo for their contribution to data analysis.\n\n\nReferences\n\nGuidance for Identifying Populations at risk of mercury exposure. UNEP -DTIE Chemicals Branch. UNEP, Geneva, Switzerland. Reference Source\n\nOdumo BO, Carbonell G, Angeyo HK, et al.: Impact of gold mining associated with mercury contamination in soil, biota sediments and tailings in Kenya. Environ Sci Pollut Res Int. 2014; 21(21): 12426–35. PubMed Abstract \| Publisher Full Text\n\nBueno PC, Bellido E, Rubí JAM, et al.: Concentration and spatial variability of mercury and other heavy metals in surface soil samples of periurban waste mine tailing along a transect in the Almadén mining district (Spain). Environ Geol. 2009; 56(5): 815–24. Publisher Full Text\n\nRodríguez Martín JA, Nanos N, Miranda JC, et al.: Volcanic mercury in Pinus canariensis. Naturwissenschaften. 2013; 100(8): 739–47. PubMed Abstract \| Publisher Full Text\n\nWang SX, Zhang L, Li GH, et al.: Mercury emission and speciation of coal-fired power plants in China. Atmos Chem Phys. 2010; 10(3): 1183–92. Publisher Full Text\n\nCarbonell G, de Imperial RM, Torrijos M, et al.: Effects of municipal solid waste compost and mineral fertilizer amendments on soil properties and heavy metals distribution in maize plants (Zea mays L.). Chemosphere. 2011; 85(10): 1614–23. PubMed Abstract \| Publisher Full Text\n\nArgyraki A, Kelepertzis E: Urban soil geochemistry in Athens, Greece: The importance of local geology in controlling the distribution of potentially harmful trace elements. Sci Total Environ. 2014; 482–483: 366–77. PubMed Abstract \| Publisher Full Text\n\nRice GE, Bullock Jr, Bullock OR, et al.: Mercury Study Report to Congress Office. Fate and Transport of Mercury in the Environment. 1997; III(Air). Reference Source\n\nRice KM, Walker EM Jr, Wu M, et al.: Environmental mercury and its toxic effects. J Prev Med Public Health. 2014; 47(2): 74–83. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDursun A, Yurdakok K, Yalcin SS, et al.: Maternal risk factors associated with lead, mercury and cadmium levels in umbilical cord blood, breast milk and newborn hair. J Matern Fetal Neonatal Med. 2016; 29(6): 954–61. PubMed Abstract \| Publisher Full Text\n\nBansa DK, Awua AK, Boatin R, et al.: Cross-sectional assessment of infants' exposure to toxic metals through breast milk in a prospective cohort study of mining communities in Ghana. BMC public health. 2017; 17(1): 505. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOgola JS, Mitullah WV, Omulo MA: Impact of gold mining on the environment and human health: a case study in the Migori gold belt, Kenya. Environ Geochem Health. 2002; 24(2): 141–57. Publisher Full Text\n\nNgure V, Davies T, Kinuthia G, et al.: Concentration levels of potentially harmful elements from gold mining in Lake Victoria Region, Kenya: Environmental and health implications. J Geochem Explor. 2014; 144(Part C): 511–516. Publisher Full Text\n\nMwamuye MK, Cherutich BK, Nyamu HM: Performance of Commercial Aquaculture under the Economic Stimulus Program in Kenya. International Journal of Business and Commerce. 2012; 2(3): 1–20. Reference Source\n\nKEMFRI: Kenya’s Aquaculture brief 2017. 2017; 12. Reference Source\n\nOkumu M, Kola S, Kanja L, et al.: Extended data on the study 'Levels of Mercury in Nile tilapia (Oreochromis niloticus), water, and sediment in the gold mining belt of Migori, Kenya, and the potential ramifications to human health' by Kola et al. figshare. Dataset. 2019. http://www.doi.org/10.6084/m9.figshare.8870053.v4\n\nHill B: Humane Harvesting of Fish. 2014; 44(209563). Reference Source\n\nPerkinElmer: Atomic Spectroscopy. A Guide to Selecting the Appropriate Technique and System. 2011; 16. Reference Source\n\nUnited States Environmental Protection Agency. Epa 7471B. 1998; 1–8. Reference Source\n\nUS-EPA: Method 7471B. MERCURY IN SOLID OR SEMISOLID WASTE (MANUAL COLD-VAPOR TECHNIQUE). United States Environmental Protection Agency. 1998; Revision 2: 1–8. Reference Source\n\nCampbell LM, Osano O, Hecky RE, et al.: Mercury in fish from three rift valley lakes (Turkana, Naivasha and Baringo), Kenya, East Africa. Environ Pollut. 2003; 125(2): 281–6. PubMed Abstract \| Publisher Full Text\n\nShelton LR, Capel PD: Guidelines for collecting and processing samples of stream bed sediment for analysis of trace elements and organic contaminants for the National Water-Quality Assessment Program. Open-File Report. 1994; 20: ill 28 cm. Publisher Full Text\n\nShafer MM, Overdier JT, Hurley JP, et al.: The influence of dissolved organic carbon, suspended particulates, and hydrology on the concentration, partitioning and variability of trace metals in two contrasting Wisconsin watersheds (USA). Chem Geol. 1997; 136(1–2): 71–97. Publisher Full Text\n\nMuller G: Index of geoaccumulation in sediments of the Rhine River. Geol J. 1969; 2: 108–18.\n\nReimann C, de Caritat P: Getting More Out of the Factsheets. In: Chemical Elements in the Environment. Springer. 1998; 11–6. Publisher Full Text\n\nUnited States Environmental Protection Agency: Guidance for assessing chemical contaminant data for use in ﬁsh advisories. Volume 1: Risk assessment and fish consumption limits. 3rd edition. United States Environmental Protection Agency, Washington, DC. 2000; 1(4305): 823-B-00–008. Reference Source\n\nEPA ABD: Risk Assessment Guidance for Superfund. Volume I: Human Health Evaluation Manual (Part A). EPA/540/189/002. 1989. Reference Source\n\nChien LC, Hung TC, Choang KY, et al.: Daily intake of TBT, Cu, Zn, Cd and As for fishermen in Taiwan. Sci Total Environ. 2002; 285(1–3): 177–85. PubMed Abstract \| Publisher Full Text\n\nCopat C, Arena G, Fiore M, et al.: Heavy metals concentrations in fish and shellfish from eastern Mediterranean Sea: consumption advisories. Food Chem Toxicol. 2013; 53: 33–7. PubMed Abstract \| Publisher Full Text\n\nCopat C, Conti GO, Signorelli C, et al.: Risk assessment for metals and PAHs by mediterranean seafood. Food Nutr Sci. 2013; 4(07): 10. Publisher Full Text\n\nHosseini SM, Mirghaffari N, Sufiani NM, et al.: Risk assessment of the total mercury in Golden gray mullet (Liza aurata) from Caspian Sea. Int J Aquat Biol. 2013; 1(6): 258–65. Reference Source\n\nMarrugo-Negrete J, Benitez LN, Olivero-Verbel J: Distribution of mercury in several environmental compartments in an aquatic ecosystem impacted by gold mining in northern Colombia. Arch Environ Contam Toxicol. 2008; 55(2): 305–16. PubMed Abstract \| Publisher Full Text\n\nSvobodova Z, Lloyd R, Machova J, et al.: Water quality and fish health. Reference Source\n\nOkumu M, Kola S, Kanja L, et al.: Underlying data (2) on the study 'Levels of Mercury in Nile tilapia (Oreochromis niloticus), water, and sediment in the Migori gold mining belt, Kenya and the potential ramifications to human health by Kola et al. figshare. Dataset. 2019. http://www.doi.org/10.6084/m9.figshare.8869964.v2\n\nGiblin FJ, Massaro EJ: Pharmacodynamics of methyl mercury in the rainbow trout (Salmo gairdneri): tissue uptake, distribution and excretion. Toxicol Appl Pharmacol. 1973; 24(1): 81–91. PubMed Abstract \| Publisher Full Text\n\nWHO: Methylmercury in Environmental Health Criteria 101. World Health Organization. 1990; 1–148. Reference Source\n\nCampbell L, Dixon DG, Hecky RE: A review of mercury in Lake Victoria, East Africa: implications for human and ecosystem health. J Toxicol Environ Health B Crit Rev. 2003; 6(4): 325–56. PubMed Abstract \| Publisher Full Text\n\nGrieb TM, Bowie GL, Driscoll CT, et al.: Factors affecting mercury accumulation in fish in the upper michigan peninsula. Environ Toxicol Chem. 1990; 9(7): 919–30. Publisher Full Text\n\nJoint FAO, Additives WHOEC on F: Summary and conclusions of sixty-first meeting. 2003.\n\nGupta A, Rai DK, Pandey RS, et al.: Analysis of some heavy metals in the riverine water, sediments and fish from river Ganges at Allahabad. Environ Monit Assess. 2009; 157(1–4): 449–58. PubMed Abstract \| Publisher Full Text\n\nEvans DW, Dodoo DK, Hanson PJ: Trace element concentrations in fish livers: implications of variations with fish size in pollution monitoring. Mar Pollut Bull. 1993; 26(6): 329–34. Publisher Full Text\n\nAsare-Donkor NK, Adimado AA: Influence of mining related activities on levels of mercury in water, sediment and fish from the Ankobra and Tano River basins in South Western Ghana. Environmental Systems Research. 2016; 5(1): 5. Publisher Full Text\n\nLamborg CH, Fitzgerald WF, Skoog A, et al.: The abundance and source of mercury-binding organic ligands in Long Island Sound. Mar Chem. 2004; 90(1–4): 151–63. Publisher Full Text\n\nPickhardt PC, Stepanova M, Fisher NS: Contrasting uptake routes and tissue distributions of inorganic and methylmercury in mosquitofish (Gambusia affinis) and redear sunfish (Lepomis microlophus). Environ Toxicol Chem. 2006; 25(8): 2132–42. PubMed Abstract \| Publisher Full Text\n\nWang WX, Wong RSK: Bioaccumulation kinetics and exposure pathways of inorganic mercury and methylmercury in a marine fish, the sweetlips Plectorhinchus gibbosus. Mar Ecol Prog Ser. 2003; 261: 257–68. Publisher Full Text\n\nBoudou A, Ribeyre F: Contamination of aquatic biocenoses by mercury compounds: an experimental ecotoxicological approach. Advances in Environmental Science and Technology (USA). 1983.\n\nDijkstra JA, Buckman KL, Ward D, et al.: Experimental and natural warming elevates mercury concentrations in estuarine fish. PLoS One. 2013; 8(3): e58401. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSnyder DM, Parker CR, Kiyoko Y: Total mercury concentration in fish tissue relative to length and weight. 2014; (Usepa 1998): 1–4. Reference Source\n\nSedláèková L, Jarkovský J, Kalina J, et al.: A negative correlation between mercury content in muscle and body weight in carp from uncontaminated ponds. Czech J Food Sci. 2015; 33(3): 204–9. Publisher Full Text\n\nOkumu M, Kola S, Kanja L, et al.: Underlying data (1) of the study 'Levels of Mercury in Nile tilapia (Oreochromis niloticus), water, and sediment in the Migori gold mining belt, Migori, Kenya, and the potential ramifications to human health by Kola et al. figshare. Dataset. 2019. http://www.doi.org/10.6084/m9.figshare.8869967.v3" }	[ { "id": "52986", "date": "03 Sep 2019", "name": "B. K. Kolita Kamal Jinadasa", "expertise": [ "Reviewer Expertise Food chemistry", "toxicology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe conclusion and abstract ending are contradictory. Ex: In the abstract, the authors said, “but is safe for the general population”. However, in conclusion, they said “there is a significant health risk associated with mercury exposure through the consumption of Nile tilapia within the Migori gold mining belt”. Moreover, in the result section, the Hg concentration of fish is more than 0.5 mg/kg, that means it is above the maximum allowable limits. So, what basis authors concluded that it is safe for general population. Hence, my suggestion is it is necessary to re-wording the abstract and conclusion part.\n\nCRM: in the results section, the authors mentioned that they run two types of CRM and spiking sample. If so, they need to point out the given value, achieved value, recovery%, number of samples, like information from the beginning of results section. But I didn’t see those CRM results under the section.\n\nHealth risk assessment. I guess, that the results are misleading to the population and scientific community. Because, you said that MS=232 g, but those are not in your country. That consumption rate was from Iran. Hence, you need to calculate Kenyan fish consume volume. Moreover, 232 g is total fish volume, the liver and brain volume (weight) were below than that. I guessed, that it will be less than 10 g. In your calculation, you assume, the whole consume volume (232 g) is equal to brain or liver (or in another word, Kenyan people consume 232 g of fish brain/liver). Hence, you got high value for THQ. So, you need to conduct basic survey to find the fish consumption rate in study area. Or if it is not practical now, you need to use Kenyan fish consumption data to calculate the health risk assessment. Meantime, authors need to observe the cooking process also, why I said that, as an example, in my country, fish liver (small fish) is going to waste and not consume.\n\nConclusion: Authors need to address the above point, and agree to accept the manuscript with major revision.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "56988", "date": "16 Dec 2019", "name": "Paul Ouboter", "expertise": [ "Reviewer Expertise Aquatic ecology", "mercury pollution." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript is much longer than needed. The section on health effects could be much shorter by just citing work by other authors. It is not really relevant for this paper. Also the chapter on \"Analytical procedures\" could be shortened by citing. Only when methods are different from what is cited, there is a need to go into full detail.\nSeveral of the tables could be made more reader-friendly by changing these into graphs.\nIn the Discussion, the authors present quite definite conclusions based on the results. It should be realized that the results are based on a very limited number of fish per site (5).\nIn the Introduction, the authors state that they have investigated farmed and wild-caught Nile tilapias. In the results the wild-caught tilapias seem to be missing.\nIn the Results, a significant difference is three times indicated as p = 0.00. This is a very unusual way to indicate significance. Better p < 0.001.\nThe relationship between fish culture and mercury levels is not treated well enough. In the Discussion the authors suggest that multi-species fish culture causes stress, thereby increasing mercury levels in the fish. The different types of fish culture were never mentioned. Maybe this information is in the Extended data. Since it is in the results and discussion, the relevant information on fish culture should be included in the main paper.\n\"Risk-based consumption limits\": Table 3 should be changed in Table 4.\nIn the discussion the authors state: \"The observation of low levels of total mercury in water sampled from fishponds may.....\". In the results they call mercury levels in water \"significantly greater than the critical value of 0.0001.g/ml\". A contradiction!\nThe suggestion that pH impacts \"distribution of mercury to the brain tissue as a receiver organ......\" is a bit far-fetched. pH has an enormous impact on the water quality, for instance on the carbonate balance. I would look for an explanation in that direction.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-1244
https://f1000research.com/articles/8-1221/v1	30 Jul 19	{ "type": "Opinion Article", "title": "Introduction to Genomic Analysis Workshop: A catalyst for engaging life-science researchers in high throughput analysis", "authors": [ "Phillip A. Richmond", "Wyeth W. Wasserman", "Phillip A. Richmond" ], "abstract": "Researchers in the life sciences are increasingly faced with the task of obtaining compute resources and training to analyze large, high-throughput technology generated datasets. As demand for compute resources has grown, high performance computing (HPC) systems have been implemented by research organizations and international consortiums to support academic researchers. However, life science researchers lack effective time-of-need training resources for utilization of these systems. Current training options have drawbacks that inhibit the effective training of researchers without experience in computational analysis. We identified the need for flexible, centrally-organized, easily accessible, interactive, and compute resource specific training for academic HPC use. In our delivery of a modular workshop series, we provided foundational training to a group of researchers in a coordinated manner, allowing them to further pursue additional training and analysis on compute resources available to them. Efficacy measures indicate that the material was effectively delivered to a broad audience in a short time period, including both virtual and on-site students. The practical approach to catalyze academic HPC use is amenable to diverse systems worldwide.", "keywords": [ "genomics", "education", "genome analysis", "high throughput computing", "hpc", "life sciences" ], "content": "Introduction\n\nThe era of genomics and DNA sequencing is being rapidly incorporated into life science research fields, spanning fields as diverse as population-scale human genetics, model organism studies and patient-focused precision medicine. Researchers have harnessed the wealth of data produced to unveil previously unattainable insights into their research questions. Although researchers across different fields are beginning to utilize the power of genomics, two major roadblocks of accessible compute resources and lack of training prevent them from being able to effectively analyze their own data1. An emerging solution to deliver high performance computing (HPC) to researchers worldwide comes through centralization of compute resources, generally in the form of grid or cluster compute servers. Examples include Compute Canada which delivers coordinated grid computing to Canadian academic institutions, the National Computational Infrastructure that coordinates the supercomputer system Raijin for Australian researchers, XCEDE which coordinates compute power for academic researchers in science and engineering across the United States, and numerous existing European academic HPC solutions such as Partnership for Advanced Computing in Europe (PRACE). Although such academic HPC systems have been historically used by researchers in physics and chemistry—fields dominated by high throughput calculations and big data computations—the systems are being increasingly used by life scientists. Even as funding for academic HPC systems grows, a lack of facilitation to introduce life scientists to their use remains a roadblock for many potential users.\n\nThe increasingly digital and quantitative analyses required in life science research has been noted for decades, but the arrival of accessible and affordable DNA sequencing technologies has accelerated the demand for skills that are not yet commonly incorporated into training programs. Thus, beyond the roadblock of HPC access and use, life science researchers must also acquire training in genomic analysis. Recent global surveys on the training needs for life science researchers in bioinformatics analysis revealed that most training comes at time-of-need or point-of-need, imposed by the necessity to analyze acquired data, instead of being a core part of the curriculum or formal education2,3. The bioinformatics community has reacted to the challenge of meeting this time-of-need training on numerous fronts with great efficacy. Current training options in bioinformatics include massive open online courses (MOOCs)4, static online tutorials5,6 and in-person workshops3,7,8. Although online forums provide a variety of resources, current tutorials are not compute-resource specific and require compute environment tailoring for HPC systems which is prohibitive to researchers lacking strong computational skills. Furthermore, surveys regarding efficacy place a high value on the practical analysis components, which are often not delivered in the online capacity due to difficulty of coordination1. In-person workshops allow for hands on applications, but require overhead cost and scheduling, which can be a limitation to many researchers. Also, many of these workshops are primarily tailored for analysis on a laptop or desktop, which differ from the environment of HPC platforms9. Moreover, some of the most popular workshops have transitioned away from introductory content, instead focusing on more complex topics and applications. Such workshops often assume background knowledge and experience in Linux and HPC. This transition occurred in part due to an expansion in online curriculum for introductory content, as well as the need to refine hundreds of workshop applicants to a feasible number of local attendees8,9. Lastly, most available workshops, both online and in-person, are rigid in structure, not catering to the interdisciplinary and diverse skill levels prevalent in the genomic era. In summary, researchers in the life sciences are faced with the need for acquiring introductory analysis skills in an environment that has the capacity for high throughput analysis, and no current training options are singular and effective at delivering this training in a time-sensitive manner.\n\nIn light of the challenges, we sought to create a new educational approach to catalyze the use of academic HPC systems by life scientists. We envisioned a flexible, centrally-organized, easily accessible, interactive, and compute resource specific training for the foundational skills of genomic analysis. To this end, we implemented and taught a modular workshop series titled: “Introduction to Genomic Analysis”. The material was taught in both an online (using the Vidyo system) and local (at the BC Children’s Hospital Research Institute) capacity for a total of seven two-hour interactive sessions. We focused the content on practical application, effective use of available compute resources, and data analysis exercises, while avoiding other aspects of genomics such as theory and experimental design. Since these aspects of genomics are fundamental, we encourage students to utilize external open-source materials deposited in curated bioinformatics education repositories10. The modularity of our workshop structure allows participants to pick-and-choose the sessions to attend based on prior experience level, thereby catering to both those new to the command-line and those experienced in Linux with an interest in exposure to genomic data analysis. Moreover, a strong emphasis was placed upon student evaluation in our workshop delivery. To benchmark student progress, problem sets were used as module exit and entrance requisites and a culminating exam assessed the ability to effectively analyze next generation DNA sequencing data.\n\nIn total, 80 participants attended at least one of the seven modules, and 58 certificates of completion were awarded based on completion of the core modules. Our initial cohort of students was diverse, including a spectrum of prior genomic analysis experience, equal gender representation, and various educational levels which recapitulates the world-wide audience4. Successful completion of the workshop showed similar results for in-person and virtual attendees and across levels of prior experience. Post-workshop surveys of the course efficacy provided deeper insight into potential improvements for future implementations of this material, and expansion into more detailed non-introductory topics of analysis.\n\nIn summary, we demonstrate the utility of our workshop format and information delivery methodology with the hopes that other trainers across the globe will improve and adapt its content to fit the needs of the life-science research community. The materials will persist in an open source state, and future implementations of the workshop will be delivered as we continue to fill the gap in genomic analysis training.\n\nIn accordance with expectations and guidelines for effective bioinformatics workshops, the materials are all open source, stored online, and video-recorded for future use11. All materials are published under Creative Commons ShareAlike 3.0 Unported License (CC BY-SA 3.0) and can be accessed at: https://phillip-a-richmond.github.io/Introduction-to-Genomic-Analysis/.\n\n\nImplementation\n\nThe materials for this workshop were designed to emphasize accessibility, consistency, modularity, and reusability, designed around an environment with capacity for high throughput analysis. We also place an emphasis on student evaluation by implementing assigned problem sets and a final exam.\n\nAccessibility. The primary challenge for delivery was the presence of both local and virtual attendees, which allowed us to expand our attendance beyond typical locally-constrained workshops. We reached both audiences simultaneously by delivering the content through an audio-visual casting of the primary teacher’s screen, which showed both lecture slides and an open terminal for executing commands. Local attendees followed the session on a projector and had tables for their laptops, while virtual attendees tuned in via internet broadcast of the screen capture using Vidyo software. Teaching assistants (TAs) monitored both local and virtual participants, the later using openstack Etherpad, and responded to questions and provided assistance during the follow-along lecture (a lecture format in which students repeat commands as they are performed by the instructor). The Etherpad environment provides an online text document updated in real-time that contains links to resources, an attendance section, and a question-and-answer section monitored by the TAs. Lectures were recorded and made available after the session for participants who couldn’t attend or desired to re-watch the presentation.\n\nConsistency. It was a goal in developing the materials to provide a consistent structure and process for each workshop session. Every session started with 45 minutes of a follow-along lecture that integrated commands for the audience to execute alongside the primary teacher. Commands executed were documented as GitHub Gists, which also contained additional details regarding command usage. At the end of the lecture, students spent 1 hour working through a problem set in small (2-3 person) groups with assistance from a roaming TA (virtually via Etherpad or locally in-person). An important introductory concept for genomic analysis is maintaining a well-organized hierarchical filesystem structure. To enforce this practice, which includes centralization of reference genomes as well as separation of raw from processed data, each session followed an identical file structure (Extended data: Supplemental Figure 1). Within this structure, individual student directories titled with unique identifiers allowed for both tracking of student participation, and simultaneous use of common files throughout the session.\n\nModularity. Modularity was a key component of this workshop as attendees had different prior training and experience. To enforce modularity and allow participants to skip material they had previously mastered, we designed each workshop session with a prerequisite assignment that assessed the comprehension of the preceding material. The problem set at the end of each session would serve as the prerequisite entry to the next. We found this to be necessary due to a mixing of content between basic Linux command-line usage and applied short-read sequencing analysis. A final exam was performed after 4 core modules, which served as a comprehensive evaluation of basic skills necessary for genomic analysis in the HPC environment. Completion of the exam was necessary for attendance of the final three modules, which delved into more advanced topics.\n\nReusability. All course materials are available through open source licensing under Creative Commons ShareAlike 3.0 Unported License (CC BY-SA 3.0), and a single github-based website links together the relevant resources. These resources include the lecture slides, problem sets, Github Gists, course exam, Etherpad links, and recordings of the lecture and problem-set sessions. Additionally, the workshop directory on the HPC environment remains for future individual use within the structured environment.\n\nHPC environment. Contrary to numerous laptop-based learning modules and teaching practices, we designed our material around analysis within a HPC environment. The motivation was two-fold: 1) by teaching in this environment we could combine an introduction to computing and Linux with an introduction to genomic analysis; and 2) we prepared researchers to be comfortable with data analysis on the platform upon which they would analyze data in the future. We utilized a national Canadian grid HPC system, Compute Canada, set up with a module system for controlling software dependencies, running Torque-Moab scheduling software for distributing jobs from the head node to the compute nodes. With slight modifications, the material can be adapted for use on most academic HPC systems. We provided temporary guest accounts to participants lacking an account. This allowed us to expose the attendees to the environment and actively engaged them to utilize the resources that are made available to them as academic researchers. Numerous participants followed up with the systems administrators to acquire full accounts during the workshop and after its conclusion.\n\nStandards in workshop design and implementation have highlighted the importance of labelling learning objectives and prerequisites explicitly for each session11. An overview of the workshop materials is displayed in Table 1. Since the workshop is modular, with benchmarks of understanding and analysis capabilities before and after each session, there is little redundancy in the per-session learning objectives. With this workshop design, we build on commands and topics from previous sessions with increasing complexity. For example, in Session 3, students learn about mapping short read DNA sequencing data to the human genome, and then, in Session 4, they re-run the same mapping commands but with the addition of variant calling. For students who are unable to master the prior material, they are encouraged to follow along and revisit the material from previous sessions. The workshop materials are available online through a coordinated website available at https://phillip-a-richmond.github.io/Introduction-to-Genomic-Analysis/.\n\nBreakdown of the workshop materials, including the learning objectives, commands learned, and prerequisites for each session.\n\n\nEvaluation of efficacy\n\nThe group of participants in the workshop series was diverse in prior training, sex, academic standing, and mode of attendance (Figure 1A–C, Figure 2A). The academic standing of our participants was similar in distribution to that reported in global surveys of researchers interested in bioinformatics related training4. The primary audience was graduate students actively completing their respective degrees, followed by undergraduates and post docs. Two faculty members attended the workshop, demonstrating the broad distribution in academic standing. Participants were evenly split between local and virtual (Compute Canada Vidyo screencast) attendees. This even distribution allowed us to draw conclusions regarding the efficacy of the teaching model where both local and virtual audiences coexist. Additionally, we had even splits in both prior genomic analysis experience and familiarity with Linux and HPC systems. Lastly, we had an equal representation of both sexes, which is an important factor in bringing equality to the field of computational biology, a historically male-dominated field12.\n\nDescription of workshop attendees including A) distribution of sexes; B) educational level; and C) the university from which they participated.\n\nA breakdown of the workshop results including A) distribution of course completion rates annotated by mode of attendance and prior experience; B) efficacy of each module based on survey responses; C) per-session attendance and problem set completion; and D) course efficacy breakdown including problem set and examination utility. Values were tallied based on attendance sign in sheets, user-submitted assignments, server workshop directories, and survey responses (Underlying data).\n\nWorkshop efficacy was determined based on three measures: 1) the number of total participants versus the number of participants that completed the exam; 2) the per-session attendance and their ability to complete the problem set; and 3) a set of questions given in a post-workshop anonymous survey. Regarding course completion, 80 attendees participated in the workshop, 59 (73.8%) of which completed the exam. The completion rate was slightly higher for the local audience (80%) than the virtual audience (66%), possibly due to the stimulation of in-person collaboration between students which was noted as lacking in the post-workshop surveys from some virtual attendees (Figure 2A). When comparing between participants with prior and “zero” experience, the completion rates were surprisingly similar (Figure 2A), with both exceeding 70%. This represents a key success, as a key intention of the novel workshop design was engagement of researchers in the life sciences with diverse levels of experience.\n\nIn analyzing the per-session attendance and problem set completion, we observed the efficacy of the module-based teaching methodology. While attendance was a measure of the interest in the subject matter, completion of the problem set identified student ability to effectively reuse the material taught during the first portion of the lesson on a unique data set. Sessions 1 through 3 had similar attendance and completion rates, and session 4 had a slightly lower completion rate since the problem set was also the mid-series exam (Figure 2B and C). The exam covered material from session 1–4, and was more in depth and difficult than preceding problem sets. Attendance for the last three sessions was optional and based on participant interest which was reflected in the lower number of attendees. Problem sets for sessions 5 and 6 were not required for attendance of the following sessions, and therefore had a lower completion rate. There was no problem set for Session 7.\n\nLastly, we analyzed the survey responses from 25 attendees for both per-session evaluations and aggregated opinions about the utility of the content. Sessions 1 through 4 are the core modules that guided attendees through an introduction to Linux and the command-line, interacting with the scheduler and queue, basics of next generation sequencing (NGS) short-read mapping and visualization, and variant calling. These sessions scored well in both efficacy and attendance, and received favorable reviews (Figure 2B). Session 5 through 7, advertised prior to the workshop as “optional”, went into more depth on subjects beyond the core NGS analysis including human genome variant interpretation and RNA-seq analysis. Despite an increase in the complexity of subject matter, the material was still judged to be accessible by the attendees. Some responses from the open commentary section were critical of the last few sessions, and suggested that those more complex topics need more time than allotted within a single 2-hour session. The progress assessments, including both problem sets and exam, were deemed useful by the majority of those that responded to the survey and received positive commentary throughout the workshop (Figure 2D). Lastly, 24/25 survey completers found the workshop useful and indicated an intent to utilize the materials and skills they learned in their own research projects.\n\n\nFuture perspectives\n\nConstructive criticism of the workshop from attendees primarily focused on the final three modules, where advanced content was presented. In future iterations of the workshop, the initial 4 core modules will be taught as a set, while advanced topics will be offered separately. Those attendees that requested longer in-person sessions were referred to offerings from consortiums, such as Bioinformatics.ca, GOBLET, and ELIXIR, each of which provides excellent advanced in-person all-day workshops. Future iterations will test whether the advanced topics are viable to be taught in the practical skills focused format highlighted in this report.\n\nOn a more granular level, the problem sets were a key part of many student’s learning, but the system for delivering and grading those problem sets, via email and posting to locations on the shared server, was inadequate. Transitioning into a web-based platform (e.g. Moodle) for assignment delivery, completion, and grading, will allow for better feedback and communication regarding problem set questions.\n\nIf you don’t use it, you lose it. After being introduced to new concepts, a new language, and new compute environments, it is critical for continued practice to maintain and refine what you have learned. As a part of our workshop design, we can contact participants in the future to see how they progress in leveraging both the HPC resources and training received to process and analyze their own datasets. Our design of the workshop around the use of centralized compute resources allows us to engage with participants to see how effective the resources are for their research purposes. These long-term metrics will help inform retention and efficacy of materials, and identify gaps in training or resource needs that we can address through partnership with the academic HPC providers.\n\nWe will deliver more workshops on both introductory and more advanced genomic analysis in the same modular format described above. By collecting similar efficacy metrics we can test how this workshop format performs with less introductory topics, and as part of a continued series. The overall goal is to establish training materials that can be delivered at time-of-need, that build strong foundations in genomic analysis utilizing the HPC systems. It is yet to be determined what the total attendee capacity is for this format, but our initial delivery reached 80 attendees, beyond the 30–40 person capacity of local workshops, and with the dual capacity of local and virtual delivery modes we anticipate audiences of over 100–200 participants.\n\n\nConclusion\n\nTraining in bioinformatics and genomics will continue to be a critical component of the development of researchers in the biological sciences. Leaders in the research domain have proclaimed that acquiring computational analysis skills should be considered on par with learning the fundamentals of wet lab techniques13. Currently, the lack of formal education in genomics and bioinformatics analysis for life-science results in numerous researchers seeking out time-of-need training to answer their research questions2.\n\nWe have introduced a practical approach to the training of life scientists that focuses on getting researchers actively engaged with an academic HPC environment available to them for continued use beyond the confines of the workshop. The format incorporates both virtual and on-site participation, and the implementation successfully enables students with varying levels of experience to engage with the training at the relevant stages. The materials are available for re-use and are adaptable for use with most academic HPC architectures (see Data availability). As more centralized HPC systems become utilized and funded, we anticipate that this format of workshop will be invaluable in providing the foundation of training for researchers in the life sciences. Upon that foundation, researchers can further specialize their training needs and effectively participate in the high throughput technology revolution.\n\n\nEthical approval\n\nFor the inclusion of attendee data and survey responses within thismanuscript, we received ethics approval from the Department of Medical Genetics at the University of British Columbia (application id H17-01298 – Student opinions of Genomic Bioinformatics Workshop). Approval was granted by Dr. Marco A. Marra, head of Medical Genetics at the University of British Columbia.\n\n\nData availability\n\nZenodo: Extended data for Manuscript: Introduction to Genomic Analysis Workshop: A catalyst for engaging life-science researchers in high throughput analysis, http://doi.org/10.5281/zenodo.334180014.\n\nThis project contains the following underlying data:\n\n- Attendee data redacted\n\n- Attendee survey responses\n\nZenodo: Extended data for Manuscript: Introduction to Genomic Analysis Workshop: A catalyst for engaging life-science researchers in high throughput analysis, http://doi.org/10.5281/zenodo.334180014.\n\nThis project contains the following extended data:\n\n- Supplementary Figure 1. Workshop Directory Structure. The workshop directory structure used for each of the sessions.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).", "appendix": "Grant information\n\nThe work was supported by NSERC Discovery Grant (RGPIN-2017-06824) acquired by WWW, and PAR was supported by a BC Children’s Hospital Research Institute Graduate Studentship award during this work.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nA previous version of this work is available from bioRxiv: https://doi.org/10.1101/478024.\n\nWe would also like to acknowledge Analise Hofmann for her assistance with ethics application and survey construction, Jana Makar of WestGrid for help organizing the workshop, Compute Canada for supporting the workshop with HPC server access and guest accounts, and the BC Children’s Hospital Research Institute Evidence2Innovation Theme for on-site coordination and support.\n\n\nReferences\n\nBrazas MD, Blackford S, Attwood TK: Training: Plug gap in essential bioinformatics skills. Nature. 2017; 544(7649): 161. PubMed Abstract \| Publisher Full Text\n\nAttwood TK, Blackford S, Brazas MD, et al.: A global perspective on evolving bioinformatics and data science training needs. Brief Bioinform. 2019; 20(2): 398–404. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrazas MD, Ouellette BF: Continuing Education Workshops in Bioinformatics Positively Impact Research and Careers. PLoS Comput Biol. 2016; 12(6): e1004916. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDing Y, Wang M, He Y, et al.: \"Bioinformatics: introduction and methods,\" a bilingual Massive Open Online Course (MOOC) as a new example for global bioinformatics education. PLoS Comput Biol. 2014; 10(12): e1003955. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSearls DB: A new online computational biology curriculum. PLoS Comput Biol. 2014; 10(6): e1003662. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSearls DB: Ten simple rules for online learning. PLoS Comput Biol. 2012; 8(9): e1002631. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStefan MI, Gutlerner JL, Born RT, et al.: The quantitative methods boot camp: teaching quantitative thinking and computing skills to graduate students in the life sciences. PLoS Comput Biol. 2015; 11(4): e1004208. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCarvalho BS, Rustici G: The challenges of delivering bioinformatics training in the analysis of high-throughput data. Brief Bioinform. 2013; 14(5): 538–47. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrazas MD, Ouellette BF: Navigating the changing learning landscape: perspective from bioinformatics.ca. Brief Bioinform. 2013; 14(5): 556–62. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAttwood TK, Bongcam-Rudloff E, Brazas ME, et al.: Correction: GOBLET: The Global Organisation for Bioinformatics Learning, Education and Training. PLoS Comput Biol. 2015; 11(5): e1004281. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchiffthaler B, Kostadima M; NGS Trainer Consortium, et al.: Training in High-Throughput Sequencing: Common Guidelines to Enable Material Sharing, Dissemination, and Reusability. PLoS Comput Biol. 2016; 12(6): e1004937. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBonham KS, Stefan MI: Women are underrepresented in computational biology: An analysis of the scholarly literature in biology, computer science and computational biology. PLoS Comput Biol. 2017; 13(10): e1005134. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEddy SR: “Antedisciplinary” science. PLoS Comput Biol. 2005; 1(1): e6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRichmond PA, Wasserman W: Extended data for Manuscript: Introduction to Genomic Analysis Workshop: A catalyst for engaging life-science researchers in high throughput analysis [Data set]. Zenodo. 2019. http://www.doi.org/10.5281/zenodo.3341800" }	[ { "id": "53165", "date": "24 Sep 2019", "name": "Richard Fitzpatrick", "expertise": [ "Reviewer Expertise Computational neuroscience", "design and evaluation quantitative and computational education for life scientists", "learning analytics" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nBiomedical research relies increasingly on the use of high-performance computing (HPC), but there is a gap in effective and timely HPC-training for practising biomedical researchers. This article describes the implementation and deployment of a modular HPC training programme. It is useful in two ways: First, the authors show that a thoughtfully designed and delivered course can achieve the goal of providing useful training on HPC that life scientists can apply to their own research. Second, by making all course materials available under a Creative Commons license, the authors have created an invaluable resource for instructors, HPC facility coordinators and self-directed learners worldwide. We also appreciated that the author share the full (de-identified) data from the participant survey.\n\nWe do feel that the article category \"Opinion Article\" does not do this work justice, since it describes a rigorously designed educational intervention and collection of resources, as well as presenting evaluation data. We leave it up to the editors to find a better suited category.\nMinor points:\nThe abstract should include information about the outcomes of the course.\n\nThe very first sentence (\"The era of genomics and DNA sequencing is being rapidly incorporated into life science research fields ...\") should be re-phrased for clarity.\n\nFigure 1: A reader's interpretation of a pie chart depends on their ability to interpret areas on the pie chart as representing the underlying numbers. Making the pie chart three-dimensional and tilting it does not add any information. Worse yet, it distorts the information available.\n\nConsider using \"gender\" instead of sex, since this is very probably the information that has been collected from participants.\n\nIn describing how the course works, we would have liked a bit more information about the teaching assistants, in particular what exactly their role was, how many there were, and how they were trained and prepared for the task.\n\nThe response rate on the post-course survey was not bad for this type of survey, but there is still a sizable proportion of non-respondents. There should be a short discussion on how this may affect survey outcomes. There is some indication (e.g. Bacon et al., Marketing Education Review, 20161) that responders are typically at the more \"extreme\" ends of the student satisfaction spectrum, i.e. students who really liked or really disliked the course are most likely to respond. As a consequence, a higher response rate would lower the scores for classes with high scores and raise the scores for classes with low scores. But on the other hand, see Nowell et al. (International Review of Economics Education, 20142) who tried to estimate non-response bias in online teaching evaluations and found it negligibly small. It may be useful for the reader to have at least an acknowledgement of the problem in the article.\n\nReference 10 is to an erratum on a paper, not the paper itself. The erratum here is important, because it clarifies the first author's last name, but we feel that the original paper should also be referenced.\n\nThe github.io site for the course is really nice, but we notice that the links to final recordings are sometimes broken and just take the user back to the main site.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes", "responses": [] }, { "id": "53168", "date": "30 Sep 2019", "name": "Michael Springer", "expertise": [ "Reviewer Expertise Systems biology" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThere is a huge need for teaching resources in computational areas for life scientist. While there are a growing number of resources, it is far from saturated. The authors focus on an important sub-area, high performance computing with an eye towards genomic sequence analysis. This is a smart choice as sequencing data has become an increasingly common tool in the life science and students recognize the need to be able to hand such data.\nThe authors deliver a course consisting of seven workshops, four core and three optional. All the resources are shared online. The results of the course and feedback is shared making the strength and areas of potential improvements clear.\nAs someone who delivers similar content to graduate student, and who has been design a similar course, I found the online resources valuable and something that I could use to build on for my own class.\nAs such, I believe the article is ready for indexing.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1221
https://f1000research.com/articles/8-920/v1	21 Jun 19	{ "type": "Research Article", "title": "Association between umbilical cord hygiene and neonatal sepsis among neonates presenting to a primary care facility in Nairobi County, Kenya: a case-control study", "authors": [ "Phoebe K. Moraa", "Marshal M. Mweu", "Peter K. Njoroge", "Marshal M. Mweu", "Peter K. Njoroge" ], "abstract": "Background: Three-quarters of all annual neonatal deaths in developing countries are attributable to neonatal sepsis. In primary care settings, poor cord hygiene due to improper handling of the infant’s cord is a major contributor to the occurrence of neonatal sepsis. The objective of this study was to describe the umbilical cord practices among mothers attending a primary care facility, assess the relationship between umbilical cord hygiene and neonatal sepsis, its impact on the population, as well as the influence of other neonatal and maternal factors on this relationship. Methods: A case-control study was conducted to assess the umbilical cord hygiene-neonatal sepsis relationship among neonates attending a primary care facility between August and October 2018. All cases were selected, while controls were systematically random sampled, as per study eligibility criteria. Exposure variables were summarized using descriptive statistics. A multivariable logistic regression model was fitted to evaluate the association between umbilical cord hygiene and neonatal sepsis adjusting for the effect of potential confounders. Subsequently, a population attributable fraction (PAF) was estimated. Results: The proportion of mothers with improper hygiene was 35.3%: 72.1% among the cases and 16.3% among the controls’ caregivers. The odds of neonatal sepsis were 13 times higher (OR=13.24; 95% CI: [7.5; 23.4]) among infants whose caregivers had improper hygiene compared to those who had proper hygiene. None of the neonatal and maternal covariates confounded the umbilical cord hygiene-neonatal sepsis association. This odds ratio gave a PAF of 66.7% (95% CI: 62.5; 69.03). Conclusions: Improper cord hygiene is prevalent in this low resource setting. Improper cord hygiene has a strong positive association with neonatal sepsis. Observing good cord care practices could avert up to 67% of newborn infections. This calls for inclusion of good cord care practices in the antenatal care educational package.", "keywords": [ "Neonatal sepsis", "Umbilical cord hygiene", "Cord care practices", "Case-control study", "Primary care setting" ], "content": "Introduction\n\nWorldwide, neonatal mortality (death occurring within the first 28 days of life) accounted for 45.1% of all child deaths in 2015, representing a 15% increase over a span of 15 years1. The leading causes of neonatal mortality globally are preterm birth complications, intrapartum-related events and neonatal sepsis1,2. These three constitute 75% of all neonatal deaths3,4. In the developing world, septicemia accounts for 1.6 million neonatal deaths per year2,5 and around 10-30% of neonatal deaths in Kenya6.\n\nOwing to the non-specificity of neonatal sepsis’ presentation in neonates, there has been a general lack of consensus on the definition of neonatal sepsis7. Nevertheless, Shane et al.8 define neonatal sepsis as a bacterial, fungal or viral systemic condition characterized by bio-physiological changes (e.g. abnormal leucocyte count, aberrant temperature or even tachycardia), clinical symptoms (e.g. presence of fever, feeding difficulties or umbilical discharge) and attended by significant morbidity and mortality8.\n\nAlthough maternal and neonatal factors are important risk factors for neonatal sepsis, umbilical cord hygiene represents a key determinant9–11. A hygienic umbilical cord refers to a dry umbilical stump without signs of redness, warmth, swelling, pain, foul smell or pus12,13. To maintain a hygienic cord, proper umbilical care is necessary. Appropriate care could be achieved by either applying methylated spirit to the base of the cord, air drying the cord to allow for natural healing or sponge-bathing neonates without immersing them in water14,15. The World Health Organization (WHO) recommends that dry cord care be employed within health facilities or home deliveries taking place in low mortality settings (less than 30 deaths per 1000 births). Chlorhexidine is advocated for home births within high neonatal mortality settings, particularly, as a substitute for harmful traditional compounds16.\n\nThe probability for entry of pathogenic micro-organisms through the umbilical cord is high in low-resource settings17,18. This could be attributable to the prevailing sub-optimal hygienic conditions in the environment of the baby that could result in a localized umbilical cord infection (omphalitis)19, with potential spread of the microorganisms into the bloodstream via the patent umbilical vessels resulting in septicemia or infection of other organs20. Although clean birth practices are highly advocated for because of their role in averting the risk of omphalitis and neonatal infection, in many developing settings, cultural norms that dictate cord care practices may compromise cord hygiene11,21. In Kenya, as in other developing settings, the rationale for applying a wide variety of substances on the cord is to hasten cord separation and healing14,17,22,23. These substances, which include cow dung, charcoal, hot fermentation, mustard oil, ghee, ash or other non-septic applications, are significantly correlated with an increased risk of omphalitis and neonatal sepsis9,10,24.\n\nIn Kenya, with a neonatal mortality rate of 22 deaths per 1000 births6, there is a lack of clear guidelines from the Ministry of Health on good cord care practices12. This lack of clarity may predispose mothers to employ suboptimal cord care practices that could lead to omphalitis and thus neonatal sepsis. Despite the importance of proper cord care in the prevention of neonatal infection, review of published literature reveals a dearth of studies that demonstrate the association between umbilical cord hygiene and systemic infection especially in poor settings15,24,25; with a sizeable number of studies paying attention to other factors associated with neonatal sepsis9,25–27.\n\nThe objective of this study was to describe the umbilical cord practices among mothers attending a primary health care facility, assess the relationship between umbilical cord hygiene and neonatal sepsis, its impact on the population, as well as the influence of other neonatal and maternal factors on this relationship. Given the absence of specific guidelines on cord care practices in Kenya, a critical understanding of the significance of good cord care on prevention of neonatal sepsis is central to informing decisions aimed at developing national guidelines on appropriate cord care practices as part of primary prevention strategies.\n\n\nMethods\n\nA facility-based unmatched case-control study design was employed to identify the determinants of neonatal sepsis. The rationale for the choice of the design relates to the rarity of neonatal sepsis within the facility’s neonatal catchment population, thus rendering the health centre a ready source of case patients. Although population-based controls would conceivably be more preferable, potential differences in health-seeking behavior between hospital and population-sourced controls suggested the need to recruit controls from the same facility as cases. The study conformed to the STROBE guidelines for reporting of a case-control study28.\n\nThe study was conducted at the Kahawa Health Centre (KHC) which is a level three state-run facility in the northern part of Nairobi County. The estimated catchment population for this health centre is about 52,193 persons and includes the adjacent peri-urban localities. Most of these areas are predominantly informal settlements characterized by overcrowding, with 99% of inhabitants being young adults. Anecdotal reports connote high neonatal mortality rates in this area.\n\nThe study population comprised all neonates presenting to KHC for pediatric services during the span of August–October 2018. Case and control patients were selected from this population based on a predefined set of eligibility criteria. All primary visit neonates (incident cases) and infants whose guardians had consented to participation were included. Premature babies with gestational age less than 37 weeks, babies who had a lower than 2000 g birth weight and neonates with congenital anomalies were excluded from the study.\n\nA case patient was a 0-28 day-old neonate, a resident of the study area, presenting to KHC during the study period with an elevated axillary temperature of ≥37.5°C and any one of the following symptoms of infection: purulent discharge (from ear/eye/umbilicus), respiratory distress/fast breathing (more than 60 breathes/minute), severe abdominal distension, poor feeding (persistent vomiting (last three feeds)/refusal to feed/inability to suck/weak suck), altered mentation (lethargic/unconsciousness/convulsions) or skin changes (deep jaundice/ periumbilical redness)29. Considering that KHC registers around two to three neonatal sepsis cases per day, to attain the computed sample, all cases (who met the aforementioned eligibility criteria) presenting to the facility within the study period were prospectively recruited. Recruitment of cases occurred at pediatric outpatient consultation rooms.\n\nControls were neonates similarly defined as cases (though devoid of sepsis symptoms), presenting to the well-baby clinic during the same two-month time period. Controls were systematically randomly sampled from the well-baby clinic of the facility frequency-matched to the cases by time of presentation.\n\nA definition for umbilical cord hygiene was adopted from WHO’s “five cleans” for postnatal care of the stump20, based on indicators that comprised of method of folding the napkin, rooming-in, bathing, handwashing and substance application practices, as reported by caregivers. An aggregate score equal to or above the median would constitute good cord hygiene, while scores below the median would be considered as poor hygiene.\n\nAs specified by Kelsey et al.30 for case-control studies, the required sample size was derived:\n\n\n\n\n\n\n\n\n\n\n\n\n\nWhere: n1 = the number of cases; n2 = the number of controls; p1 = the proportion of cases with an unhygienic umbilical cord; p2= proportion of controls with an unhygienic umbilical cord specified at 37.6% based on a previous study31. Notably, Zα/2 = 1.96 for the 2-tailed confidence level of 95%; Z1-β = −0.84 for the desired statistical power of the study set at 80%; and r=2 as the specified ratio of controls to cases. The odds ratio (OR) for the umbilical cord hygiene-neonatal sepsis association was estimated at 2. With an anticipated 5% non-response rate, the required sample size was 312: 104 cases and 208 controls.\n\nOther than the primary exposure variable, the other predictor variables were maternal and neonatal factors. These were gathered using a semi-structured questionnaire, available as Extended data32. Maternal factors consisted of socio-demographic factors (age of mother, level of education, marital status, parity and religion), the antenatal history of the mother (number of antenatal care (ANC) visits, history of receiving health education, tetanus toxoid immunization, prenatal maternal bacterial infection, birth attendance, place of delivery and type of delivery) and post-natal history factors (history of illness or pregnancy related complications such as postpartum depression, nutritional status or other comorbidities). Neonatal risk factors included low APGAR scores of <7 at 5 minutes (whose signs included scores of appearance, pulse, grimace, activity, and respiration), neonate’s age, sex and invasive procedures (use of medically invasive instruments/resuscitation at birth). Table 1 displays assessment of these variables. Figure 1 provides a conceptual framework of the relationship between the aforementioned predictors and the outcome.\n\nThe research commenced after receiving written clearance from the Kenyatta National Hospital (KNH)-University of Nairobi (UoN) Ethics and Research Committee (P438/06/2018) and the Nairobi County Health Services (Ref. No. CMO/NRB/OPR/VOL.1/2018/91). Additionally, written informed consent was obtained from the mother/index care-giver for their neonate’s participation in the study.\n\nPrior to commencing the data collection, two research assistants were trained on screening patients, complete neonatal medical examination and standardized interview techniques to reduce interviewer bias. Additionally, caregivers could have resorted to unhygienic cord practices (such as harmful applications) as a way to treat an already septic neonate. Hence, the possibility of reverse causality was reduced by focusing on incident cases. Attempts to minimize recall bias were made by referencing the mother-child booklet to ascertain information regarding some antenatal and perinatal information such as the number of antenatal visits, the neonate’s date of birth, neonatal APGAR score and resuscitation history.\n\nThe questionnaires were checked for completeness and qualitative data coded. The data were double-entered by two independent data entry clerks into EpiData version 3.1 spreadsheet. The principal researcher cross-checked the computerized data base against the questionnaires that had been administered. The dataset was exported to Stata software, version 13 (Stata Corporation, College Station, Texas, USA) for cleaning and analysis. For continuous variables’ descriptive statistics, data were summarized by means, medians and ranges. For categorical variables, data were summarized using frequency tables, proportions and percentages.\n\nScoring of the umbilical cord hygiene variable’s five components was standardized such that those responses that were desirable as per WHO essential newborn care guidelines received a higher value (≥1). A value of zero was awarded to responses inconsistent with these guidelines41–43. A total score was then reached by summing up the individual component’s scores. Notably, a cord with an aggregate score below the median score of 7 was designated as improper cord hygiene, whereas one with a score equal to or above the median was deemed to have proper cord hygiene.\n\nA logistic regression model was used to assess the crude association between umbilical cord hygiene and neonatal sepsis. For a sensible interpretation of APGAR score’s effect on neonatal sepsis, it was grouped into two categories27: ≥7 or <7. To evaluate the potential confounding effect of neonatal and maternal factors on the umbilical cord hygiene-sepsis relationship, each of the predictors was screened for unconditional associations with neonatal sepsis at a 5% significance level. Qualifying variables were further screened for a significant association with umbilical cord hygiene at similar level of significance.\n\nVariables that met these criteria were considered as potential confounders to the cord hygiene-neonatal sepsis relationship and therefore were included in a multivariable model to adjust for their confounding effect on this relationship. Here, a backward step-wise approach was applied to eliminate variables if there was not more than a 30% change in the regression coefficient for umbilical cord hygiene upon their exclusion44. To evaluate the impact of umbilical cord hygiene in the neonatal population (the proportion of neonatal sepsis that could be prevented by adhering to proper umbilical cord care), a PAF was computed as described by Dohoo et al.44\n\n\n\nWhere: AFp is the Population Attributable Fraction; pd is the proportion of total cases in the population arising from improper cord hygiene; aOR is the adjusted odds ratio for cord hygiene derived from the multivariable model.\n\n\nResults\n\nA total of 312 participants (104 cases, 208 controls) were recruited into the study but those who consented to participation were 309. Of the 208 potential controls, three declined consent. Additionally, three others did not meet the eligibility criteria and were excluded, leaving 202 eligible controls that participated. A flow diagram illustrating the recruitment and enrollment process is shown in Figure 2.\n\nDescriptive statistics for the demographic variables are indicated in Table 2. Notably, males comprised 55.8% (n=58) of cases and 47.0% (n=95) of controls. The mean neonatal age was 19.7 days; the mean age of cases and controls being 16.5 days (range: 5-28 days) and 21.3 days (range: 3-28 days), respectively. Regarding marital status, 69.2% (n=72) of cases’ mothers were married compared to 81.7% (n=165) of controls’. Only 13.5% (n=14) of the cases’ caregivers had received up to tertiary level of education compared to 22.3% (n=45) of the controls’.\n\nA description of the participants’ cord care practices is displayed in Table 3. In this population, majority of mothers reported use of chlorhexidine/surgical spirit (64%, n=197). Among cases, slightly over a third (35.6%, n=37) had surgical spirit/chlorhexidine applied as compared to about four-fifths (79.2%, n=160) of the controls. Of concern, saliva/ash was applied among 10.6% (n=11) of cases compared to 2.5% (n=5) of the controls. In this study setting, about two-thirds (65.7%, n=201) of mothers fastened their babies’ diapers below the umbilical stump. Roughly 30% (29.8%, n=31) of the case respondents revealed that they folded the neonate’s napkin below the cord in comparison to 84% (84.2%, n=170) of the control participants.\n\nAlmost four-fifths (78.1%) of caregivers declared that they washed their hands when changing the napkins/diapers. This consisted of 54.8% (n=57) of the cases’ and 90.1% (n=182) of the controls’ mothers. Regarding the cleansing substance employed by those who reported handwashing, only 44.4% (n=136) used both water and soap. In particular, whereas 61.4% (n=124) of controls’ mothers stated they used water and soap before cord handling, only 11.6% (n=12) of cases’ mothers did the same. The practice of rooming-in was followed by approximately all (99.4%, n=304) of the participants. Amongst these were 98.1% (n=102) of the cases and all the controls. Sponge-bathing was the bathing practice recorded by most (64.4%, n=197) of the participants in the present study. However, only 28.9% (n=30) of the cases were sponge-bathed.\n\nThe proportion of mothers/care-givers who had improper hygiene practices was 35.3% (n=108), with unhygienic cord status being disproportionately high in case (72.1%, n=75) than in control mothers (16.3%, n=33), Table 3.\n\nThe crude association between cord hygiene and neonatal sepsis is captured in Table 4. Notably, the odds of neonatal sepsis in infants who had improper hygiene was approximately 13 times higher (OR=13.24; 95% CI: [7.5; 23.4]) compared to those with proper hygiene.\n\n* Variables eligible for an assessment of their association with the primary exposure (P≤0.05). CI, confidence interval.\n\nOf the variables screened, neonatal factors registering a significant association with neonatal sepsis were: low APGAR score (P=0.001), invasive procedures (P=0.007) and neonate’s age (P<0.001). With respect to maternal factors, marital status (P=0.05), initiation of breastfeeding (P=0.006), type of feed (P<0.001) and pregnancy-related events (P=0.005) were found to be significantly associated with neonatal sepsis (Table 4). To qualify as potential confounders, these significant factors were further evaluated for an association with the primary exposure as presented in Table 5. Following the assessment, the variables: low APGAR score, invasive procedure, neonate’s age, marital status, type of feed and pregnancy-related events were significantly associated with cord hygiene and thus were offered to the multivariable model to adjust for their potential confounding effect.\n\na,b,c,d,e,f Variables eligible for inclusion in the multivariable analysis (P≤0.05). CI, confidence interval.\n\nFrom the multivariable analysis, none of the six factors assessed confounded (resulted in a >30% change in the coefficient for umbilical cord hygiene) the primary association between umbilical cord hygiene and neonatal sepsis (Table 6), and as such, the OR of 13.34 was used to compute the PAF. The estimated PAF was 66.7% (95% CI: 62.5; 69.0).\n\nCI, confidence interval; None of the assessed factors resulted in a >30% change in the regression coefficient for umbilical cord hygiene.\n\n\nDiscussion\n\nThe study found that in this community, the main cord care procedures involved aspects of substance application, with a majority of caregivers cleansing their hands using water and soap (44%), exposing the cord (66%), sponge-bathing (64%) and practicing rooming-in (99%) of the mother-infant couplet. This is in line with WHO recommendations on satisfactory cord care in high mortality regions: use of select topical antimicrobial agents as alternatives to harmful applications, handwashing, air drying of the umbilical stump, sponge bathing and rooming-in16,20.\n\nThis study found that the most commonly used agents for treatment of the cord were chlorhexidine or surgical spirit (64%). In other studies, similar frequencies in the use of these antimicrobials as the principal cord care application substances have been reported41,45,46. However, there was a statistically significant difference between the 36% of cases whose caregivers used surgical spirit/chlorhexidine and the 79% of sepsis-free controls, highlighting the importance of surgical spirit use in prevention of sepsis. Of concern was the significant number of mothers who used non-recommended substances which included water or nothing (air-drying) and ash/saliva. Such unclean substances are a probable nidus of infection as they are likely to be contaminated with bacteria/spores24,45,47. In similar settings, variations with respect to the most popularly applied substances have been observed. For instance, in Pumwani Maternity Hospital, Kenya, applying nothing (air drying) was most prevalent at 55%, followed at 25% by surgical spirit, as well as use of saliva and water both at 10%40. Elsewhere, methylated spirit was the main cord care method in Ghana48 and in Nigeria41,46,49; while use of brick ash was reportedly highly used in Zambia22. Findings from another study carried out in Benin reported that inappropriate/harmful substances were applied by 81% of caregivers50. In Ethiopia and Nigeria, dry cord care was widely exercised51,52. The differences might be due to the influence of deeply entrenched cultural norms that supersede adoption of advocated clean cord care applications10,23,53.\n\nIn this study, about two thirds of mothers tied the babies’ diapers below the umbilical stump. This is in consonance with WHO stipulations that dictate that the diaper should be tied below the cord20. There was a clear distinction among cases and controls, with only 30% of cases and 84% of controls’ mothers reporting to fasten diapers below the cord. A study by Kinanu et al.40 found similar results where 54% neonates’ diapers were applied below the cord. The umbilical stump being an acquired wound is a nidus for entry of pathogenic bacteria from the newborn’s excreta16,29,40.\n\nAbout four-fifths (78%) of caregivers mentioned that they washed their hands while changing the diapers (55% of cases and 90% of controls). Another study done in one public hospital in Nairobi, Kenya, supported this finding where 52% of mothers washed their hands under running water and 48% used water in basins40. Comparably, in Parakou, Benin, 73% of mothers expressed that they washed their hands prior to cord care provision50. Further, with regards to the washing substance in the current study, majority of mothers (44%) used water and soap. Mothers who washed with plain water were 37% while those who did not wash their hands at all were 22%. Findings in this study were corroborated by a Nigerian study which documented most of the study population (47%) to have used water and soap in the care of their hands, followed by water only (40%)41. Nonetheless, a study in Karamoja in Uganda, contrasted the finding, reporting that handwashing was not observed by majority (90%) of mothers before change of diaper/napkin leading to their neonates exhibiting signs of infection54. The difference could be ascribed to the Ugandan study area being primarily a semi-arid region of the country compared to the urban setting of this study. While handling the neonate’s cord, it is recommended that handwashing with both water and soap is observed to achieve umbilical cord hygiene11,20.\n\nOver 99% of the mothers in this study slept in the same room as the baby. The finding from a study in Pumwani Maternity Hospital in Kenya supports this result where 93.3% of mothers were shown to practice rooming-in40. It is recommended that mothers and their newborns should sleep in one room throughout without separation16. It has been cited that rooming-in promotes better coupling of mother and newborn, boosting their skin contact and hence increasing colonization rates of non-pathogenic organisms from the mothers’ normal skin flora to the baby, thereby lowering umbilical cord infection rates14,20,55.\n\nTo achieve dry cord care and hastened healing, the bathing practice is key. Wiping the baby with a wet cloth was dominant among controls in the present study. Similarly, sponge-bathing has been shown in another study to be the main bathing practice compared to immersion-bathing40. However, majority of cases were immersed in water. In Benin, 93% of mothers tub-bathed babies in water and only 7% wiped them with a wet cloth which was linked to concomitant umbilical cord infection50. The WHO recommends that the first bath should be delayed for at least six hours and umbilical stump should be kept dry until the cord falls off20; the reason being that immersion bathing leads to delay in cord separation and increased susceptibility to sepsis56.\n\nThis study results showed that 35.3% of caregivers failed to observe good cord hygiene. In North Benin, a study reported that, as per study’s specifications of cord hygiene, 58.6% of mothers had practiced poor quality care, 31.9% had good quality care, with none of the mothers reaching excellent quality of cord care50.\n\nThe results of the present study demonstrated a statistically significant association between umbilical cord hygiene and neonatal sepsis among infants of the Kahawa Health Centre. Compared to babies whose mothers observed proper cord hygiene, the odds of developing neonatal sepsis among babies of mothers who had improper cord hygiene was roughly 13 times higher (OR=13.24; P<0.001) and this key association was not confounded by any of the examined factors. According to Bradford Hill criteria, such a strong association has been shown to be less likely due to chance, bias or confounding and might suggest causality57. However, this finding needs to be validated by studies in other settings.\n\nIn India, a previous study has elucidated a strong association (P<0.0001) between unhygienic care of the cord and sepsis24. Likewise, a study in Bangladesh showed a relative risk of 1.15 for an association between unclean cord care and neonatal sepsis58. The strength of association is lower than the results of this study perhaps attributable to other stronger predictors of neonatal sepsis in the population. A similar observation was made in Nigeria where unhygienic cord care was strongly associated with neonatal infection15.\n\nWith the strong OR, this study yielded a high overall PAF estimate of 67% for umbilical cord hygiene. This implies that in the study's neonatal population, sixty-seven percent of neonatal sepsis cases would have been averted, if good cord hygiene was observed and assuming umbilical cord hygiene was causal. Associations drawn from this study are generalizable to similar low-resource primary care settings.\n\nA few limitations are intrinsic to the present study. Recall of past exposures was likely to be more complete in respondents whose neonates were cases than controls. This could lead to biased effect estimates away from unity. Moreover, there was likely to be non-differential reporting between cases’ and controls’ caregivers disproportionately related to cord care consequently biasing the effect estimates towards null.\n\n\nConclusion\n\nThis study provides evidence that improper cord hygiene is strongly associated with neonatal sepsis among infants presenting in this primary care setting. Most importantly, this association was not confounded by any of the covariates measured. The PAF estimate implies that observance of good cord hygiene practices would result in a 67% reduction of sepsis in this neonatal population. Hence, there is a pressing need to spearhead revision of national guidelines with a view to introducing an antenatal cord care package that lays emphasis on the importance of comprehensive proper cord care practices.\n\n\nData availability\n\nHarvard Dataverse: Replication Data for: Association between umbilical cord hygiene and neonatal sepsis among neonates presenting to a primary care facility, Kenya: A case-control study. https://doi.org/10.7910/DVN/FSXPR832.\n\nThis project contains the following underlying data:\n\nkahawa_hygiene_code.do (.do file code for umbilical cord-neonatal sepsis evaluation).\n\nkahawa_hygiene_data.tab (study dataset).\n\nHarvard Dataverse: Replication Data for: Association between umbilical cord hygiene and neonatal sepsis among neonates presenting to a primary care facility, Kenya: A case-control study. https://doi.org/10.7910/DVN/FSXPR832.\n\nThis project contains the following extended data:\n\nUmbilical hygiene_sepsis questionnaire.pdf (questionnaire used in this study).\n\nData are available under the terms of the Creative Commons Zero “No rights reserved” data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nThis work was supported by the University of Nairobi scholarship (Ref No. 10007492016).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe owe our sincere gratitude to the mothers and caregivers of the Kahawa Health Centre and their infants who facilitated data collection and through whom this research work was able to come to fruition; implementation of the recommendations drawn from the study findings would be of greatest value to them. We also need to thank the administrative team and health workers of the KHC for their hospitality and allowing utilization of the facility to enable this work’s realization.\n\n\nReferences\n\nLiu L, Oza S, Hogan D, et al.: Global, regional, and national causes of under-5 mortality in 2000-15: an updated systematic analysis with implications for the Sustainable Development Goals. Lancet. 2016; 388(10063): 3027–35. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOza S, Lawn JE, Hogan DR, et al.: Neonatal cause-of-death estimates for the early and late neonatal periods for 194 countries: 2000-2013. Bull World Health Organ. 2015; 93(1): 19–28. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDivision of Policy and Strategy U: Committing to child survival: a promise renewed. progress report 2013. New York: UNICEF. 2013. Reference Source\n\nUNICEF: Committing to Child Survival: A Promise Renewed. Progress Report. 2014. Reference Source\n\nVergnano S, Sharland M, Kazembe P, et al.: Neonatal sepsis: an international perspective. Arch Dis Child Fetal Neonatal Ed. 2005; 90(3): F220–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKDHS, Kenya National Bureau of Statistics (KNBS) and ICF Marco 2015: Kenya Demographic and Health Survey. Key Indicators Report: Infant and child mortality. Calverton, Maryland: KNBS and ICF Macro, 2014. Reference Source\n\nWynn JL, Wong HR, Shanley TP, et al.: Time for a neonatal-specific consensus definition for sepsis. Pediatr Crit Care Med. 2014; 15(6): 523–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShane AL, Sánchez PJ, Stoll BJ: Neonatal sepsis. Lancet. 2017; 390(10104): 1770–80. PubMed Abstract \| Publisher Full Text\n\nMullany LC, Darmstadt GL, Katz J, et al.: Risk factors for umbilical cord infection among newborns of southern Nepal. Am J Epidemiol. 2007; 165(2): 203–11. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAmbe J, Bello M, Yahaya S, et al.: Umbilical Cord Care Practices In Konduga Local Government Area Of Borno State North - Eastern Nigeria. The Internet Journal of Tropical Medicine. 2008; 5(2): 2–6. Reference Source\n\nBlencowe H, Cousens S, Mullany LC, et al.: Clean birth and postnatal care practices to reduce neonatal deaths from sepsis and tetanus: a systematic review and Delphi estimation of mortality effect. BMC Public Health. 2011; 11 Suppl 3: S11. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMinistry of Public Health and Sanitation, Ministry of Medical Services: National Guidelines for Quality Obstetrics and Perinatal Care. MOH Kenya, 2010. Reference Source\n\nBugaje MA, McHoney M, Ameh EA, et al.: Omphalitis. Paediatric Surgery: A Comprehensive Text For Africa. 2010; 124–8. Reference Source\n\nWhitmore JM: Newborn Umbilical Cord Care: An Evidence Based Quality Improvement Project. Doctor of Nursing Practice (DNP) Projects. 2010; 13. Reference Source\n\nSaleh JA, Nemecek J, Jones C: Impact of hygienic caring of the umbilical cord in the prevention of neonatal tetanus. WebmedCentral PUBLIC HEALTH. 2015; 6(5): WMC004891. Reference Source\n\nWHO: WHO recommendations on postnatal care of the mother and newborn. World Health Organization; Geneva, 2013. Reference Source\n\nAmare Y: Umbilical cord care in Ethiopia and implications for behavioral change: a qualitative study. BMC Int Health Hum Rights. 2014; 14: 12. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAkter T, Dawson A, Sibbritt D: What impact do essential newborn care practices have on neonatal mortality in low and lower-middle income countries? Evidence from Bangladesh. J Perinatol. 2016; 36(3): 225–30. PubMed Abstract \| Publisher Full Text\n\nKarumbi J, Mulaku M, Aluvaala J, et al.: Topical umbilical cord care for prevention of infection and neonatal mortality. Pediatr Infect Dis J. 2013; 32(1): 78–83. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWHO: Care of the umbilical cord. WHO/FHE/MSM-cord care World Health Organization; Geneva, 1998. Reference Source\n\nWaiswa P, Pariyo G, Kallander K, et al.: Effect of the Uganda Newborn Study on care-seeking and care practices: a cluster-randomised controlled trial. Glob Health Action. 2015; 8: 24584. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSacks E, Moss WJ, Winch PJ, et al.: Skin, thermal and umbilical cord care practices for neonates in southern, rural Zambia: a qualitative study. BMC Pregnancy Childbirth. 2015; 15: 149. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCoffey PS, Brown SC: Umbilical cord-care practices in low- and middle-income countries: a systematic review. BMC Pregnancy Childbirth. 2017; 17(1): 68. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGoel A, Murmu SK, Shah S, et al.: Role of cultural practices in neonatal sepsis. Int J Med Sci Public Health. 2015; 4(5): 680–3. Publisher Full Text\n\nJohn B, David M, Mathias L, et al.: Risk factors and practices contributing to newborn sepsis in a rural district of Eastern Uganda, August 2013: a cross sectional study. BMC Res Notes. 2015; 8: 339. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJabiri A, Herman LW, Avelina S, et al.: Prevalence and factors associated with neonatal sepsis among neonates in Temeke and Mwananyamala Hospitals in Dar es Salaam, Tanzania. Tanzan J Health Res. 2016; 18(4): 1–7. Reference Source\n\nGebremedhin D, Berhe H, Gebrekirstos K: Risk Factors for Neonatal Sepsis in Public Hospitals of Mekelle City, North Ethiopia, 2015: Unmatched Case Control Study. PLoS One. 2016; 11(5): e0154798. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nvon Elm E, Altman DG, Egger M, et al.: The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol. 2008; 61(4): 344–9. PubMed Abstract \| Publisher Full Text\n\nWHO: Problems of the neonate and young infant. Pocket book of hospital care for children. World Health Organization, Second edition, 2013a; 51–6. Reference Source\n\nKelsey JL, Whittemore AS, Evans AS, et al.: Methods in Observational Epidemiology. 2nd ed, New York, 1996. Reference Source\n\nKinanu L, Odhiambo E, Mwaura J, et al.: Socio-demographic and economic determinants of umbilical cord infection among neonates at Pumwani maternity hospital, Kenya: a cross-sectional study. Int J Health Sci Res. 2015; 5(12): 274–81. Reference Source\n\nKeraka P: Replication Data for: Association between umbilical cord hygiene and neonatal sepsis among neonates presenting to a primary care facility, Kenya: A case-control study. Harvard Dataverse, V1. 2019. http://www.doi.org/10.7910/DVN/FSXPR8\n\nKhan R, Vandelaer J, Yakubu A, et al.: Maternal and neonatal tetanus elimination: from protecting women and newborns to protecting all. Int J Womens Health. 2015; 7: 171–80. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDalili H, Nili F, Sheikh M, et al.: Comparison of the four proposed Apgar scoring systems in the assessment of birth asphyxia and adverse early neurologic outcomes. PLoS One. 2015; 10(3): e0122116. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRazaz N, Boyce WT, Brownell M, et al.: Five-minute Apgar score as a marker for developmental vulnerability at 5 years of age. Arch Dis Child Fetal Neonatal Ed. 2016; 101(2): F114–20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOishi T, Iwata S, Nonoyama M, et al.: Double-blind comparative study on the care of the neonatal umbilical cord using 80% ethanol with or without chlorhexidine. J Hosp Infect. 2004; 58(1): 34–7. PubMed Abstract \| Publisher Full Text\n\nShwe DD, Abok II, Diala UM, et al.: Methylated spirit versus 4% chlorhexidine gel in neonatal umbilical cord infection: A short report of a randomized, open-labelled, parallel-group trial. Niger J Paediatr. 2018; 45(2): 118–22. Reference Source\n\nStewart D, Benitz W, COMMITTEE ON FETUS AND NEWBORN: Umbilical Cord Care in the Newborn Infant. Pediatrics. 2016; 138(3): pii: e20162149. PubMed Abstract \| Publisher Full Text\n\nImdad A, Bautista RM, Senen KA, et al.: Umbilical cord antiseptics for preventing sepsis and death among newborns. Cochrane Database Syst Rev. 2013: (5): CD008635. PubMed Abstract \| Publisher Full Text\n\nKinanu L, Odhiambo E, Mwaura J, et al.: Cord Care Practices and Omphalitis among Neonates Aged 3 - 28 Days at Pumwani Maternity Hospital, Kenya. J Biosci Med. 2016; 04(1): 27–36. Publisher Full Text\n\nAfolaranmi TO, Hassan ZI, Akinyemi OO, et al.: Cord Care Practices: A Perspective of Contemporary African Setting. Front Public Health. 2018; 6: 10. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAmolo L, Irimu G, Njai D: Knowledge of postnatal mothers on essential newborn care practices at the Kenyatta National Hospital: a cross sectional study. Pan Afr Med J. 2017; 28: 97. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTeklehaimanot HD, Teklehaimanot A, Tedella AA, et al.: Use of Balanced Scorecard Methodology for Performance Measurement of the Health Extension Program in Ethiopia. Am J Trop Med Hyg. 2016; 94(5): 1157–69. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDohoo I, Martin W, Stryhn H: Methods in Epidemiologic Research. Prince Edward Island: Charlottetown, 2012. Reference Source\n\nAbhulimhen-Iyoha BI, Ofili A, Ibadin MO: Cord care practices among mothers attending immunization clinic at the University of Benin Teaching Hospital, Benin City. Niger J Paediatr. 2011; 38(3): 104–8. Publisher Full Text\n\nAbegunde D, Orobaton N, Beal K, et al.: Trends in newborn umbilical cord care practices in Sokoto and Bauchi States of Nigeria: the where, who, how, what and the ubiquitous role of traditional birth attendants: a lot quality assurance sampling survey. BMC Pregnancy Childbirth. 2017; 17(1): 368. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPeterside O, Duru CO, Anene N: Harmful traditional practices in a newborn: A case report. Niger J Paediatr. 2015; 42(2): 151–3. Reference Source\n\nNutor JJ, Kayingo G, Bell JF, et al.: Knowledge, attitudes and practices regarding care of newborn umbilical cord among healthcare workers and mothers in the Volta region of Ghana. Ann Global Health. 2016; 82(3): 548. Publisher Full Text\n\nOpara PI, Jaja T, Dotimi DA, et al.: Newborn Cord Care Practices Amongst Mothers in Yenagoa Local Government Rea, Bayelsa State, Nigeria. Int J Clin Med. 2012; 03(1): 22–7. Publisher Full Text\n\nAgossou J, Hounnou-d’Almeïda M, Adédémy JD, et al.: Newborn Umbilical Cord Care in Parakou in 2013: Practices and Risks. Open J Pediatr. 2016; 06(1): 124–35. Publisher Full Text\n\nCallaghan-Koru JA, Seifu A, Tholandi M, et al.: Newborn care practices at home and in health facilities in 4 regions of Ethiopia. BMC Pediatr. 2013; 13: 198. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChidiebere O, Uchenna E, Stanley O, et al.: Umbilical Cord Care Practices and Incidence of Febrile Illnesses in the First Month of Life among Newborns- A Population Based Study. Br J Med Med Res. 2015; 5(11): 1422–30. Publisher Full Text\n\nWalsh S, Norr K, Sankar G, et al.: Newborn cord care practices in Haiti. Glob Public Health. 2015; 10(9): 1107–17. PubMed Abstract \| Publisher Full Text\n\nHopp LJ: Delivery practices, hygiene, birth attendance and neonatal infections in Karamoja, Uganda: a community-based study. Afr Health Sci. 2017; 17(1): 7–13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMacDonald L: Becoming Baby Friendly: Rooming-in for Patient Centered Care in the Maternal Setting. Honors College Theses. 2016; Paper 21. Reference Source\n\nAyyildiz T, Kulakci H, Niyazi Ayoglu F, et al.: The effects of two bathing methods on the time of separation of umbilical cord in term babies in Turkey. Iran Red Crescent Med J. 2015; 17(1): e19053. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCox LA Jr: Modernizing the Bradford Hill criteria for assessing causal relationships in observational data. Crit Rev Toxicol. 2018; 1–31. PubMed Abstract \| Publisher Full Text\n\nMitra DK, Mullany LC, Harrison M, et al.: Incidence and risk factors of neonatal infections in a rural Bangladeshi population: a community-based prospective study. J Health Popul Nutr. 2018; 37(1): 6. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "50237", "date": "08 Jul 2019", "name": "Balafama Alex-Hart", "expertise": [ "Reviewer Expertise Neonatology." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper describes umbilical cord practices among mothers in a low resource setting and assesses the relationship between umbilical cord hygiene and neonatal sepsis, its impact on the population, as well as the influence of other neonatal and maternal factors on this relationship. The methodology is clearly written and well presented. However, criteria used for cases were purely clinical and do not quite fit into the seven clinical signs identified by WHO especially in developing countries which includes: difficulty feeding, convulsions, movement only when stimulated, respiratory rate >60 per min, severe chest in drawing and axillary temperature >37.5 °C or <35.5 °C1. For example, the authors include an axillary temperature of >37.5 °C but are silent on temperature < 35.5 °C. Other useful signs like cyanosis and grunting are also not included2. It is important to remember that neonatal sepsis shares similar clinical presentations to other common conditions in the neonatal period, thus a combination of clinical and laboratory findings is necessary to provide a correct diagnosis of neonatal sepsis. The limited access to laboratory tests in developing countries 3 which perhaps may have been a challenge to the authors should probably be stated if applicable.\nThe study describes the statistical methods that were used and the results obtained are clearly stated. Statistical analysis and interpretation of results are adequate. This is a very good paper and is an important addition to the literature on the importance of cord care especially in low resource settings.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "50236", "date": "16 Jul 2019", "name": "Jamlick Karumbi", "expertise": [ "Reviewer Expertise Epidemiology." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an important article that clearly describes what may be going wrong in our quest to reduce neonatal mortality in high mortality settings like Kenya. The article does point out that the correct umbilical cord care practices need to be entrenched into routine practices, especially in lower-level facilities.\n\nThere are a few issues that need to be clarified;\nIn the introduction, the authors state that there are no guidelines for cord care in Kenya, contrary to that there are actually basic pediatric protocols of 2016 here. These protocols clearly recommend the use of 4% Chlorhexidine for cord care.\n\nThe authors used a 1:2 case-control matching, what was the rationale? Why not 1:1? My guess is that this has to do with powering the study enough, if this is the case then it needs to be clearly stated in the methodology.\n\nIn the results section, Table 3 seems to be entirely described by the text preceding it. I believe the authors can just highlight the key elements in the table and leave the other details in the table.\n\nThe authors could have enriched the discussion by contextualizing the findings further, for example, commodities are usually a challenge in low-level facilities and also human resources who would have correctly advised the mothers, I was hoping to see this as possible postulations in the discussions and conclusions.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-920
https://f1000research.com/articles/8-357/v1	01 Apr 19	{ "type": "Method Article", "title": "A bilayer tissue culture model of the bovine alveolus", "authors": [ "Diane Lee", "Mark Chambers", "Mark Chambers" ], "abstract": "The epithelial lining of the lung is often the first point of interaction between the host and inhaled pathogens, allergens and medications. Epithelial cells are therefore the main focus of studies which aim to shed light on host-pathogen interactions, to dissect the mechanisms of local host immunity and study toxicology. If these studies are not to be conducted exclusively in vivo, it is imperative that in vitro models are developed with a high in vitro-in vivo correlation. We describe here a co-culture bilayer model of the bovine alveolus, designed to overcome some of the limitations encountered with mono-culture and live animal models. Our system includes bovine pulmonary arterial endothelial cells (BPAECs) seeded onto a permeable membrane in 24 well Transwell format. The BPAECs are overlaid with immortalised bovine alveolar type II epithelial cells and the bilayer cultured at air-liquid interface for 14 days before use; in our case to study host-mycobacterial interactions. Characterisation of novel cell lines and the bilayer model have provided compelling evidence that immortalised bovine alveolar type II cells are an authentic substitute for primary alveolar type II cells and their culture as a bilayer in conjunction with BPAECs provides a physiologically relevant in vitro model of the bovine alveolus.\n\nThe bilayer model may be used to study dynamic intracellular and extracellular host-pathogen interactions, using proteomics, genomics, live cell imaging, in-cell ELISA and confocal microscopy. The model presented in this article enables other researchers to establish an in vitro model of the bovine alveolus that is easy to set up, malleable and serves as a comparable alternative to in vivo models, whilst allowing study of early host-pathogen interactions, currently not feasible in vivo. The model therefore achieves one of the 3Rs objectives in that it replaces the use of animals in research of bovine respiratory diseases.", "keywords": [ "Replacement", "NC3Rs", "bovine", "alveolus", "type II", "endothelial", "air-liquid interface" ], "content": "\n\nRemoves the requirement to conduct multiple in vivo studies.\n\nImmortalisation of cells allows consistency of data between studies.\n\nEasy set-up allows studies which cannot feasibly be performed in live animal models.\n\nAllows first-line investigations of specific disease pathways in vitro, hence is suited to drug discovery screens and toxicology studies.\n\nProvides an alternative to in vivo experiments on bovine respiratory diseases, negating mild-moderate invasive procedures in up to 100 cows per year (UK) used in bovine respiratory studies, approximately 40 of which will be used in BTB vaccine studies (Annual Statistics of Scientific Procedures on Living Animals for Great Britain, 2017).\n\nEnables users to obtain representative alveolar bilayers with desirable physiological and structural features of the native bovine alveolus within 3 weeks, rather than months, as in the case of in vivo studies.\n\nImmortalised cell lines facilitate reproducibility and comparisons with previously published literature.\n\nCost effective compared to performing similar experiments in vivo.\n\nEnables researchers without animal facilities to study bovine respiratory diseases.\n\nThe immortalised cell lines are suitable for studies involving monocultures of alveolar type II cells, whilst the bilayer may be used for transport studies of candidate pathogens, drug molecules, vaccines and interactions thereof.\n\nCurrently used by researchers to study early host-pathogen interactions of Mycobacterium bovis and the live attenuated vaccine Bacillus Calmette–Guérin (BCG) in conjunction with peripheral blood mononuclear cells.\n\nThe bilayer model could potentially incorporate bovine macrophages and also be used to investigate the role of surfactant proteins in innate immunity.\n\nSets the standard to recapitulate structural and functional features for future in vitro models of the alveolus.\n\nThe bilayer alveolus model may be translatable to other species, with minor modifications.\n\n\nIntroduction\n\nThe lung is a complex organ, lined in its entirety by specialised epithelium. The most distal region of the lung consists of the alveoli, designed primarily for efficient gas exchange and arranged in clusters or acini1. The alveolar epithelium is composed of two types of cell; alveolar type I (ATI) and alveolar type II (ATII). Alveolar type I cells cover 90 % of the alveolar surface2, despite only constituting in the region of 7 % of the epithelium by numbers3. This can be attributed to the elongated squamous cell morphology of the ATI, which spreads over a large surface area and lies in close proximity to capillary endothelial cells. This lends the ATI to its primary role of gas exchange and enables the regulation of physiological solute transport between the alveolus and the circulatory system4. Previous studies have provided evidence for the role for ATI cells in fluid homeostasis of the alveolar compartment, with the finding that sodium ions are transported via the epithelial sodium channel (ENaC), using active ion transport5.\n\nThe ATI cell is terminally differentiated, being derived from the alveolar type II (ATII) cell. Unlike the former, the ATII cell is compact and ‘cuboidal’. Early studies performed in simians and rodents generated evidence to suggest that ATII cells expressing surfactant protein C are the main progenitor cells of the alveolar epithelium6–8. This role was cemented by the findings of later research, including lineage tracing studies9 and morphological characterisation10. Proliferation of the ATII cell is normally relatively slow compared to other epithelial cells9; however ATII hyperplasia has been reported in response to injury, such as that inflicted by the chemotherapeutic agent bleomycin, providing further evidence that ATII cells are the main progenitor of the lung11.\n\nThe close proximity of alveolar type I epithelial cells to the neighbouring capillary endothelial cells forms a highly gas permeable bilayer barrier across which gas exchange occurs. The structural integrity of the alveolus is maintained during its expansion and contraction by the formation and continuous secretion of a phospholipid-rich film (pulmonary surfactant) from ATII cells that spreads and covers the alveolar epithelial cell surface.\n\nPerhaps most pertinent to the study of host-pathogen interactions is the role of ATII cells in innate immunity, protecting the lungs against respiratory infection. The ATII cell was first coined as the ‘defender of the alveolus’ by Mason and Williams in 197712, following numerous studies into the phospholipid content of the surfactant secreted by primary ATII. As with mucus of the upper airways, the composition of surfactant varies according to disease status13,14. In particular, surfactant protein D (SP-D) has been hallmarked as a biomarker for pulmonary disease on account of its variability when quantified from bronchoalveolar lavage (BAL) or serum15. Along with surfactant protein A (SP-A), SP-D has been found to bind to viruses, bacteria, yeast and fungi. Both act as opsonins, thus increasing the pathogen’s susceptibility to phagocytosis16–19.\n\nIt follows that further study of host-pathogen interactions of pathogens with ATII cells is critical to advances in the field. Costs, low throughput and ethical concerns are all limitations of in vivo models. In addition, early events at the host-pathogen interface cannot be studied easily in vivo, due to the transient nature of such early events. These limitations may be overcome by in vitro systems; however, few systems exist that accurately reproduce the structural complexity of the in vivo environment, particularly for non-human species. A human bilayer model of the alveolus was first described in 199920 and was used to study early events in Mycobacterium tuberculosis infection. However, no equivalent in vitro bilayer model of the bovine alveolus is available to enable comparative studies between species, according to a search of the PubMed database using the terms, ‘bovine’, ‘alveolus’, ‘bilayer’ and ‘model’ (November 2018). In vivo models of M. bovis and M. tuberculosis infection result in ‘non-recovery’ of subjects and thus raise strong ethical considerations21. Cattle are natural hosts of M. bovis and are therefore one of the target species (alongside wildlife reservoirs such as badgers and wild boar) for strategies aimed at the study and control of bovine TB. It follows that the most relevant host-pathogen interaction studies would be performed by experimental infection of cattle. Furthermore, the close relationship of M. bovis to M. tuberculosis21 gives rise to synergy in the efforts to find novel vaccines and therapeutic interventions to control human and bovine TB. For example, vaccines which were developed for use in humans have been evaluated in the bovine M. bovis live animal model with reported success, whilst immune responses and reaction to therapeutic strategies are also comparable22–24. The relatively large size of cattle, however, presents a major challenge. They require particularly specialist husbandry and housing at biosafety level 3 containment, which counter the advantages of frequent and sequential sample collection and infringe upon the ‘Five Freedoms’, as defined by the Farm Animal Welfare Council25.\n\nContinuous epithelial cell lines offer the advantage of reduced inter-experiment variability, extended proliferative capacity and better suitability for high-throughput screening. Such a cell line was generated from bovine ATII cells by Su et al. (2013)26. However, upon enquiry, this was found to be no longer available. We have therefore developed a simple method of isolation for ATII cells27 and generated two new cell lines, the bovine alveolar type II (BATII) and the bovine type 2 alveolar epithelial (B2AE) cell lines. These ATII derived cells have been integrated into an in vitro tissue culture bilayer model, with bovine pulmonary arterial endothelial cells (BPAECs). Cultured at an air-liquid interface, this bilayer recreates the fundamental elements of the bovine pulmonary alveolus.\n\nOur model advances the replacement of cattle by providing a readily available and reproducible means to study fundamental events following infection of the bovine lung with virulent pathogens that cannot be conducted currently in vitro. The model is also amenable to translational and applied research, screening and toxicity studies, especially involving zoonotic pathogens, such as M. bovis and M. tuberculosis, or respiratory syncytial virus, the human and bovine strains of which are closely related. The ease of set up and small footprint per experiment also lends the model well to the generation of pharmacological data for novel therapies targeted at respiratory disease or topical agents delivered via inhalation. This applies to diseases such as bovine respiratory disease (BRD), for which new antimicrobial therapies are urgently needed in order to develop therapeutic strategies and perform pharmacokinetic studies28. In the UK alone, the Annual Statistics of Scientific Procedures on Living Animals (2016) reports that a total of 95 procedures were performed on live cattle for respiratory research29. We can, therefore, propose with confidence that upwards of this number could potentially be replaced, per year, globally.\n\n\nMethods\n\nThe isolation procedure is outlined in detail by Lee et al. (2018)30. All efforts were made to ameliorate animal suffering. This was achieved in the current study by performing the isolation procedure on tissue sourced from freshly slaughtered cattle less than 24 months of age at a local abattoir facility, under existing licensed slaughter procedures. Details of the reagents required for isolation and culture of ATII cells may be found in Table 1. A 5 cm3 piece of distal right lung was taken for disease surveillance. Tissue was dissected into pieces no greater than 3 mm3 and washed in Dulbecco’s Phosphate Buffered Saline (DPBS) containing 100 U/mL each of penicillin and streptomycin, with five repetitions, with a further two washes in DPBS containing the same concentration of penicillin and streptomycin as well as 2 mM EDTA. Digestion was carried out for 30 minutes at 37°C in DPBS containing 100 U/mL each of penicillin and streptomycin, 0.008 % (w/v) elastase (Fisher Scientific, UK), 0.2 % (w/v) collagenase (Fisher Scientific), 0.005 % (v/v) DNAse Type I (2000 KU/mL; Sigma, St.Louis, MI, USA) and 0.05 % (w/v) trypsin (Gibco). Enzymatic activity was neutralised with an equal volume of Dulbecco’s modified Eagle Medium/Ham’s F12 (DMEM/F12) (Gibco) containing 25 % FBS (heat-inactivated, Gibco) and 0.01 % (v/v) DNAse I. The cell suspension was filtered sequentially through 100, 50 and 25 μm filters (Biodesign™, distributed through Fisher Scientific) and spun at 300 × g for 10 minutes at room temperature, before resuspending in 1:1 DMEM/F12/Small Airway Growth Medium (SAGM) (Lonza, UK), 5 % FBS and 0.025 % (v/v) DNAse I.\n\nThe cell suspension was overlaid onto petri dishes coated with bovine IgG (5 μg/mL; Sigma Aldrich), for 1 hour at 37°C, rocking after 30 minutes to redistribute non-adhered cells. Non-adherent cells were removed with the supernatant and spun at 300 × g for 5 minutes. The resulting pellet was resuspended in 4 % Percoll™ solution (Sigma Aldrich) and overlaid onto a gradient consisting of 30 % ‘heavy’ Percoll™ and 10 % ‘light’ Percoll™. Gradients were spun at 400 × g for 20 minutes at 4°C. Enriched ATII cells were identified as a band at the 10–30 % interface. These were removed and washed in DPBS containing 100 U/mL each of penicillin and streptomycin, resuspended in SAGM, plated onto 6-well plates and cultured at 37°C, 5 % CO2.\n\nDetails of the reagents required for immortalisation of ATII cells may be found in Table 1. ATII cells were immortalised using lentivirus particles in two ways. The BATII cell line was generated by transduction with particles encoding the catalytic subunit of human telomerase (hTERT) and Simian Virus 40 large T antigen (SV40), whilst the B2AE utilised hTERT and the proto-oncogene B cell-specific Moloney murine leukemia virus integration site 1 (Bmi1). SV40 lentivirus was purchased as a ready to use particle suspension (AMS Biotechnology (Europe) Ltd, Abingdon, UK). To generate hTERT and Bmi1 lentiviral particles, the hTERT (human open reading frame clone pENTR221 IOH 36343) and Bmi1 (human open reading frame clone pENTR221 IOH13688) were purchased as Entry Vectors (Invitrogen Life Sciences, Carlsbad, CA, USA). The gene of interest was sub-cloned into a destination vector, in an LR Clonase™ recombination reaction using the Gateway Cloning Technology system (Invitrogen Life Sciences). This resulted in pLenti6-hTERT and pLenti6-Bmi1 expression vectors, or ‘expression clones’. The expression vectors were transformed into chemically competent One Shot Stbl3 E. coli, provided as a component of the Gateway Cloning Technology System. Insert integrity for both expression vectors was assessed by restriction enzyme digestion (XhoI) and sequencing. The latter was performed using primer sequences designed using the Primer-BLAST tool31, generating primers directed at internal and flanking sites (Table 2) in the expression vector.\n\nPrimers for the internal sequencing segments were designed using the Primer-BLAST online tool, whilst the external CMV and V5 (C-terminal) sequences were provided by Invitrogen.\n\nLive lentivirus particles were generated according to the protocols described in the Invitrogen Life Sciences ViraPower™ II Lentiviral Gateway® Expression Kit (Invitrogen Life Sciences). Briefly, 3 μg of pLenti6-hTERT or pLenti6-Bmi1 expression vector was mixed with 9 μg Virapower™ packaging mix in 1.5 mL Opti-MEM I medium (Invitrogen Life Sciences). This was added to 36 μL Lipofectamine™ 2000 transfection reagent (Invitrogen Life Sciences) (pre-diluted in 1.5 mL Opti-MEM I). The DNA-Lipofectamine complexes were added to a 10 cm tissue culture treated plate containing 5 mL of Opti-MEM I/10 % FBS. To this, 6 × 106 293FT producer cells (supplied with the kit) were added to 5 mL Opti-MEM I/10 % FBS, following Invitrogen’s reverse transfection protocol. Lentiviral particles were harvested as supernatant at 24, 48 and 72 hours post transfection and combined, before centrifugation at 300 × g for 15 minutes to pellet cell debris. The supernatant was then filtered through a 0.45 μm PVDF filter and centrifuged at 20,000 rpm, using a TH-641 rotor in a Sorvall WX80+ ultracentrifuge for 90 minutes at 4°C, to pellet viral particles.\n\nViral titre was determined using the HeLa cell line (p2 after receipt) according to the Virapower™ protocols and recommendations, using dilutions at 10-2, 10-3, 10-4, 10-5 and 10-6. HeLa cells were seeded at 2 × 105 cells/well in a 6-well plate, in DMEM/10 % FBS/1 × non-essential amino acids (‘complete medium’, all Gibco). On the day of transduction (24 hours after seeding), cell supernatant was removed and the viral supernatant dilutions, prepared in complete medium, were added directly to the cell cultures (1 mL total culture volume). Medium only was added to a ‘mock’ well, used to confirm cell death in the presence of the selection antibiotic, blasticidin. To each well, 6 μg of Polybrene® transfection reagent (Sigma Aldrich) was added. Cells were returned to the incubator and cultured at 37°C, 5 % CO2 for 24 hours. Medium containing the viral supernatant was then replaced with 2 mL complete culture medium and the plates returned to the incubator for a further 24 hours. At 48 hours after transduction, blasticidin was added to all wells except untreated controls at a concentration of 10 μg/mL. Medium containing blasticidin was replaced every 3–4 days, until 14 days, when all cells were dead in the mock well and discrete antibiotic-resistant colonies could be seen in wells transduced with viral supernatant. These were counted following visualisation with 1 % crystal violet solution (Sigma Aldrich) and titre determined by averaging the counts of two wells, following correction for dilutions. Titred virus was stored at -80°C until use.\n\nPrimary ATII cells were immortalised at p1, five days after isolation. Cells were seeded into four wells of a 24-well plate at a density of 0.5 × 105 cells/mL, 0.5 mL per well, in complete SAGM, without antibiotics. At 24 hours, SV40 and hTERT lentiviral particles were added simultaneously at a multiplicity of infection (MOI) of 10 (for each virus), in 0.5 mL SAGM, again without antibiotics. At 24 hours post infection, the media was replaced with fresh SAGM (1 mL). Cultures were fed every 2–3 days with SAGM and passaged accordingly, alongside primary ATII cells for comparison and grown without selection, monitoring morphology, proliferation rate and karyotype.\n\nGrowth curves for the immortalised cell lines were performed at passages 2, 10 and 18, comparing them with those of primary ATII cells at passage 2, using the CellTitre 96® AQueous One Proliferation Assay (Promega, Madison, WI, USA). Cell number and viability were determined by trypan blue exclusion (cells displaying uptake of trypan blue were considered non-viable) using a TC-20 automated cell counter (BioRad, Hercules, CA, USA). For each cell type/passage, cells were resuspended to a final concentration of 0.5 × 105 cells/mL in SAGM. An aliquot (100 μL, 5000 cells) was dispensed into three wells of a 96-well plate (x 9 plates), before incubation in a humidified, 5 % CO2 atmosphere at 37°C. At 24 hours, one plate was removed. To each well, 20 μL of CellTiter 96® AQueous One Solution Reagent was added, returning the plate to the incubator for 1 hour. Absorbance was then measured at 490 nm, using the CLARIOstar® 96-well plate reader (BMG Labtech, Ortenburg, Germany). The assay procedure was repeated for the remaining time points. Resulting absorbances were entered as interpolated values against a reference curve generated from seeding pre-determined cell numbers in the same format and performing the assay after a 1hour incubation period.\n\nTo determine genetic integrity and characterise the stability of the immortalised cell lines, a metaphase chromosomal spread was carried out to ascertain karyotype of the BATII cell line at passages 4, 14 and 23 and of the B2AE cell line at passages 7, 12 and 22. The preparation of each spread was based on the protocol published by Campos et al.32, with some deviations. Briefly, BATII, B2AE or wild-type ATII cells at the relevant passage were revived from liquid nitrogen stores and cultured for 48 hours in SAGM, in a humidified, 5 % CO2 atmosphere at 37°C. Cells were then treated with Colcemid™ (Sigma Aldrich) at a final concentration of 0.2 μg/mL in SAGM. Cells were returned to the incubator for 4 h, before trypsinisation with 0.25 % trypsin/0.05 % EDTA (Gibco/Invitrogen Life Sciences) for 5 minutes at 37°C. After trypsin neutralisation, cells were centrifuged at 122 × g for 5 minutes. The supernatant was discarded and 5 mL pre-warmed (37°C) hypotonic solution (KCl 75 mM(aq)) was added in a two-stage process: 3 mL was added to the tube in an inclined position, whilst rotating the tube to mix cells, followed by the remaining 2 mL to the upright tube. The cells and hypotonic solution were placed at 37°C for 15 minutes to allow swelling. To fix cells, three drops of room temperature fixative solution (freshly prepared methanol/glacial acetic acid 3:1) were added to the tube, followed by inversion to mix and centrifugation (without brake) at 122 × g for 5 minutes. The supernatant was again discarded, leaving 200 μL hypotonic solution. Fixative solution (3 mL) was added to the tube at an incline, rotating as before, followed by a further 2 mL to the tube wall, to wash attached cells down. The tube was stored upright overnight at 4°C, before centrifugation as before. Supernatant was again discarded, leaving 200 μL, and the pellet was resuspended by flicking the tube. Fixative solution (5 mL) was added as before and the centrifugation step repeated, again leaving 200 μL supernatant and resuspending by flicking the tube. An aliquot of each sample (20 μL) was spread onto a microscope slide pre-cleaned in 6M HCl, moving the pipette tip across the surface during dispensing. The slide was then passed through steam (face up). Slides were mounted with coverslips in Vectorshield Hardset with DAPI (Vector Laboratories, Burlingame, CA, USA) and examined under a fluorescence microscope to determine karyotype.\n\nTo test the effects of an extracellular matrix on the culture of ATII cells, we coated tissue culture in plastic and also permeable inserts in a 10 % solution of growth factor reduced Matrigel™ (356230, Corning). Matrigel™ was thawed on ice overnight in a cold room and kept on ice throughout the procedure. Pre-cooled SAGM was stored on ice and Matrigel™ added to a final concentration of 10 %. Using chilled pipette tips, the solution was used to coat either 12 mm diameter Transwell™-CLEAR 3.0 μm pore size PET inserts (3462, Corning, New York, US) or the surface of wells in a 24-well tissue culture plate (3337, Corning) (50 μL per cm2). Coating was performed overnight at 2–8°C and stored until use (up to 2 weeks). Coated surfaces were gently washed once with SAGM to remove excess Matrigel™.\n\nBovine pulmonary artery endothelial cells (BPAECs) were purchased from the European Collection of Authenticated Cell Cultures (ECACC) (Salisbury, UK) and cultured in endothelial growth medium (EGM-2 Bulletkit) (Lonza, UK). Cells were used between passages 3 and 10 only.\n\nA schematic of the bilayer assembly is shown in Figure 1. BPAECs were revived from liquid nitrogen and cultured to 80 % confluence in EGM-2. Cells were trypsinised using 0.25 % trypsin/0.05 % EDTA and neutralised using an equivalent volume of DMEM/10 % FBS, before seeding onto the apical surface of 6.5 mm diameter Transwell-CLEAR™ 8μm pore size permeable membranes in a 24-well plate, at a density of 2 × 104 cells/insert, (approximately 6.5 × 104 cells/cm2). EGM-2 (600 μL) was added to the basolateral chamber of each well. BPAECs were cultured for 5–7 days, replacing EGM-2 in the basolateral chamber and removing apical medium which had seeped through from the basolateral side of the membrane. BATII (or B2AE) cells were revived 3 days after BPAEC and cultured in SAGM without antibiotics. On the day of seeding, cells were trypsinised, resuspended in EGM-2 and counted, before seeding on top of the BPAEC layer at a density of 2 × 104 cells/insert, as before. For histology, additional bilayers were seeded onto 12 mm diameter Transwell-CLEAR™ 0.4 μm pore size (also 6.5 × 104 cells/cm2 for each cell line), adding 1.5 mL EGM-2 basolaterally. In each case, the bilayer was returned to the incubator for 2 hours to allow for attachment, after which the apical medium was removed and the cells cultured at air-liquid interface for 14 days, feeding every 2–3 days basolaterally (600 μL) and removing any medium on the apical surface. Monolayers were also prepared for each cell type, using the same seeding densities and culture methods/feeding intervals for comparison.\n\nBovine pulmonary arterial endothelial cells (BPAECs) are seeded onto the apical surface of the insert membrane and cultured for 5 days. The immortalised Bovine Alveolar Type II (BATII) epithelial cells are then overlaid onto the BPAECs and the bilayer cultured for 14 days as outlined in the methods section. Also shown is a timeline detailing the step(s) required at each timepoint.\n\nThe formation of tight junctions was assessed for monolayers of both cell types and bilayers in three ways. Firstly, each mono (BPAEC or BATII only) or bilayer was visually assessed, observing the apical surface by eye to estimate coverage by EGM-2 which had come through from the basolateral chamber.\n\nSecondly, trans-epithelial electrical resistance (TEER) was measured between the time of seeding until the day of harvest. TEER values were measured using an EVOM2 Voltohmmeter with STX-2 chopstick electrodes (World Precision Instruments, Stevenage, UK) immediately before the medium was exchanged. For measurements, 0.5 mL and 1.0 mL of medium were added to the apical and basolateral chambers, respectively, allowing the medium to equilibrate to 37°C before measurements were performed in triplicate. All values were converted to Ohms/cm2 using Equation 1.\n\nFinalTEER(Ω/cm2)=NetTEER(Ω)×AreaofTranswellinsert(cm2)(Equation 1)\n\nThe third method of determining layer integrity utilised the dextran permeability method, as outlined by Birkness et al. (1999)20. A 10 mg/mL solution of Dextran Blue 2000 (DB2000; Sigma Aldrich) was prepared in DPBS, alongside standard solutions of 7.5, 5.0, 3.75, 2.5, 1.25, 0.625, 0.3125 and 0.156 mg/mL. DPBS (600 μL) was added to each well of a fresh 24-well plate. Inserts containing cells were removed one by one from the original plate, removing any apical medium, and transferred to the fresh plate. DB2000 (100 μL of 10 mg/mL) was added to the apical surface of each insert during the transfer. The plate was then placed at 37°C, 5 % CO2, for 1 hour. A blank insert was treated in the same way. An empty well had 600 μL DPBS added, followed by 100 μL of 10 mg/mL DB2000 in DPBS (total volume 700 μL). This was designated ‘no membrane control’. At the end of the incubation, 600 μL was removed from the basolateral compartment and read at 630 nm on a spectrophotometer, alongside the prepared standards. An aliquot (60 μL) was taken from apical solutions and diluted in 540 μL DPBS (a ten-fold dilution), before reading at 630 nm. Unknown values were extrapolated from a linear regression of the standard curve.\n\nFor characterisation of the novel cell lines, BATII and B2AE cells were seeded onto Nunc Lab-Tek II 8 chamber coverglass slides (Thermo Fisher). Cells were seeded at 4 × 104 cells per chamber and cultured for 72 hours. Slides were then washed with DPBS and fixed in 4 % PFA at room temperature for 15 minutes, before permeabilisation by washing in an immunofluorescence wash buffer (IF buffer) (Table 3) containing 0.1 % triton X-100 for 15 minutes. Blocking was performed for 1 hour at room temperature in DPBS/ 5 % normal goat serum/ 0.1 % triton X-100. Primary antibodies were diluted in blocking buffer, applied to cells and incubated overnight at 4°C. Cells were rinsed three times with IF buffer and secondary antibody applied for 1 hour at room temperature in the dark. Cells were again rinsed three times with IF buffer and stored at 4°C until imaging, using a Nikon Eclipse Ti confocal microscope.\n\nThe following ingredients are added to ~475 mL reverse osmosis water and the resulting solution buffered to pH 7.4. Top up to 500 mL, then store at 2–8°C. Remove to room temperature before use and dilute 1:10 with reverse osmosis water.\n\nFor immunofluorescence (IF) analysis of bilayer cultures, inserts were washed with DPBS and processed as for coverglass monolayer cultures. Following the final wash to remove the secondary antibody, the membrane was removed from the insert using a scalpel blade, then mounted in Prolong® Gold Antifade reagent with 4′,6-diamidino-2-phenylindole (DAPI) (Fisher Scientific) or Vectashield Hardset mounting medium with phalloidin (TRITC) (Vector Laboratories, Burlingame, CA, US) between two cover slips. Bilayers were imaged using a Nikon Eclipse Ti confocal microscope.\n\nPrimary antibodies are detailed below in Table 4. Mouse and rabbit IgG1 isotype controls were used accordingly.\n\nFor histology, fixed membranes were dehydrated sequentially in 70, 90 and 100 % ethanol, for 30 minutes at each concentration. Inserts were then incubated in 100 % isopropanol for 30 minutes, before transfer to molten paraffin wax (65°C) for 1 hour. The membrane was then excised from the insert using a scalpel blade and embedded in paraffin, cured and sectioned at 4 μm. Sections were subjected to dewaxing and H&E staining as outlined in Lee et al.30. Stained sections were dehydrated and mounted in DPX mounting medium (Sigma Aldrich) with a cover-slip overlaid for analysis on a Nikon Eclipse Ci upright microscope.\n\nAll statistical analysis was performed using GraphPad Prism version 7.03 for Windows, GraphPad Software, La Jolla, California, USA, www.graphpad.com. Replicate numbers and number of experiments are detailed in figure legends. Dextran blue measurements (Figure 7A–B) were evaluated using a one-way ANOVA and Tukey’s multiple comparisons tests. These are presented as means ± standard deviation, where n represents individual inserts, selected at random by the researcher and assigned into groups (not blind). Statistical significance is denoted as * p ≤ 0.05, p ≤ 0.01, * p ≤ 0.001 and **** p ≤ 0.0001 (Tukey’s multiple comparisons test shown). Sample size was limited by plate size and practical considerations (handling of multiple samples of this nature). We therefore performed three repeats of each experiment (n = 3 inserts per group) on different days, using fresh cells and inserts to ensure replicability of findings. To increase the statistical power of each test, data from identical groups of each day was first analysed using ANOVA and a Tukey’s multiple comparison’s test for variability, before combining replicates from each day to give a total of 9 replicates per group. Gaussian distribution was assumed for ANOVA; however, sphericity (equal variability of differences) was not and thus, the Geisser-Greenhouse correction was used.\n\nTEER analysis was performed for the average of three readings per insert at each time point, with six inserts per group, per experiment (BATII and B2AE bilayers). Data shown is combined from three experiments performed on different days and presented as means ± standard deviation, analysed using an unpaired t-test.\n\nHere we describe the step by step procedure used to set up and validate a 14-day bilayer of BATII (ATII-derived) epithelial cells with BPAEC endothelial cells on Transwell inserts. A simplified schematic of the procedure, including timeline, is found in Figure 1, whilst all reagents may be found, with their provenance, in Table 1. In principle, the same protocol also applies to the pairing of B2AE cells in a bilayer with BPAECs; however, it should be noted that the authors have not optimised this arrangement, having chosen to focus on BATII cells for reasons outlined in the discussion.\n\nStep 1: Revive BPAEC cells (Day -9). All stages of culture are carried out in a humidified incubator, at 37°C, 5 % CO2.\n\nWe use cells between passages 2 and 10, for consistency and to eliminate the potential for phenotypic drift at higher passages. Cells, frozen down at 1 × 106/mL/vial in Recovery™ Cell Freezing Medium (12648010, Invitrogen), were revived from liquid nitrogen, seeded at a density of 2 × 106 cells per T75 cm2 flask in EGM-2 (CC-3162, Lonza, UK) and cultured to 80 % confluence, without antibiotics. EGM-2 media (500 mL) should be made up on the day of revival from liquid nitrogen, using freshly thawed Lonza bullet kit Singlequots™. After a period of one month, any remaining media of the batch should be discarded and fresh media prepared. One bottle is sufficient to maintain BPAEC cells (using 10 mL per T75 cm2 flask) and also feed one 24 well plate of bilayer cultures for a total of 21 days (including the initial culture phase, consisting of BPAEC cells only). Medium is exchanged the day after seeding (day -8) and at 2 days (day -6), always ensuring that the cells have media exchanged the day before trypsinisation.\n\nThese cells are not sensitive to penicillin and streptomycin; however, our downstream applications require the bilayer system to be antibiotic free.\n\nStep 2: Seed BPAECs onto Transwells (Day -5). BPAEC cells are trypsinised using 3 mL of 0.25 % trypsin/0.05 % EDTA (25200056, Life Technologies, UK) per T75 cm2 flask and neutralised using an equivalent volume of DMEM/10 % FBS. The suspension is triturated to ensure homogeneity of the suspension, centrifuged at 300 × g for 5 minutes, then resuspended in 1 mL EGM-2 for counting, using the TC-20 automated cell counter. This takes the form of a viability count, mixing 10 μL of cell suspension with 10 μL of 0.4 % trypan blue (15250061, Life technologies). BPAEC cells are then seeded onto the apical surface of 6.5 mm diameter Transwell-CLEAR™ 8μm pore size permeable membranes in a 24 well plate, at a density of 2 × 104 cells/insert, in 100 μL EGM-2 (approximately 6.5 × 104 cells/cm2). EGM-2 (600 μL, the volume recommended by Corning) is added to the basolateral chamber of each well. BPAECs are cultured on the Transwell inserts for 5–7 days, replacing EGM-2 in the basolateral chamber at 2-day intervals and removing apical medium which seeps through from the basolateral side of the membrane. BPAECs do not form sufficient tight junctions to prevent media seepage – this step is simply to remove old media from the apical surface.\n\nStep 3: Revival of BATII cells (Day -3). The BATII cell line has thus far not been karyotyped beyond passage 23, therefore we have restricted use between p14 and p23 in our own experiments. We freeze down BATII cells at a density of 1 × 106 cells/mL/vial in Recovery™ Cell Freezing Medium (as before). BATII cells should be revived 3 days prior to seeding onto Transwell inserts and cultured in SAGM (CC-3118, Lonza) without antibiotics, seeding 1×106 cells per T75 cm2 flask, to allow for the relatively high proliferation rate of these cells compared to BPAECs. As with EGM-2, SAGM should be prepared fresh using the Lonza Singlequots™ on the day of revival and any media remaining after one month discarded. BATII cells are fed the day after seeding into flasks and also the day before trypsinisation, at a volume of 10 mL per T72 cm2 flask.\n\nStep 4: Seeding of BATII cells onto the BPAEC layer (Day 0). On the day of seeding, cells are trypsinised and neutralised as for BPAECs, resuspended in 1 mL EGM-2 and counted, as before. A dilution is prepared in EGM-2, at a density of 2 × 105 cells/mL. Seeding on top of the BPAEC layer is performed at a density of 2 × 104 cells/insert. The bilayers are then returned to the incubator for 2 hours to allow for attachment, after which the apical medium should be removed and the cells cultured at air-liquid interface (ALI) for 14 days. Cultures achieve ALI (defined as less than 25 % of the insert submerged by EGM-2) from around 5 days after the seeding of BATII cells.\n\nStep 5: Feeding and TEER monitoring (Day 2 – 14). Cultures are maintained in a humidified incubator, at 37°C, 5 % CO2, feeding every 2–3 days in the basolateral compartment only (600 μL) and removing any media on the apical surface.\n\nTechnical tip: Media removal or addition is performed by tilting the plate to a 30° angle and placing the tip edge against the side of the well. Additions are performed slowly. This helps the user to avoid touching the cell layers and minimises cell stress.\n\nTEER studies are carried out from 2 days after the seeding of BATII cells onto BPAEC cells in Transwells, i.e. the generation of a bilayer. BPAECs do not form sufficient tight junctions compared to a bilayer and this this is reflected in the TEER values, when comparing bilayers or BATII monolayers to BPAEC monolayers. Measurements are acquired using the EVOM2 epithelial Voltohmmeter, with a 4 mm STX2 chopstick electrode (300523, World Precision Instruments Inc., US). When monitoring trans-epithelial electrical resistance (TEER), on the day of feeding, fresh EGM-2 should first be placed in the apical chamber (500 μL) as well as the basolateral compartment (600 μL). A blank Transwell insert is set up in another well for comparison. The nature of TEER measurement means that drift is to be expected, particularly as the plate cools down during measurement, therefore the medium should be equilibrated to temperature in situ (at least 15 minutes) before measurements are taken. Ideally, this temperature is 37°C; however, in practice, this is difficult to achieve if a full 24 well plate is to be measured. We therefore perform 3 measurements in each well (moving from well 1 to 24, then repeating the process twice), using chopstick electrodes which are first sterilised in 70 % ethanol and rinsed in DPBS. Net TEER is calculated by subtracting the reference (blank Transwell insert only) from the measured TEER of the bilayer or monolayer sample (Equation 2). This is then converted to Final TEER by correcting the value for the area of cell growth (Equation 3).\n\nApical medium is removed after TEER measurements, to maintain ALI. Average TEER is plotted for each well at a particular time point. As part of our characterisation studies, the ‘average of averages’ (average of 3 readings taken in each replicate well, n=3 Transwells per timepoint) for each time point were plotted between days 2 and 14 of bilayer culture, with standard deviation.\n\nTechnical tip: to keep readings as consistent as possible, hold the chopstick electrodes in the same place for each insert in a perpendicular fashion to the plate surface, with the shorter electrode in the apical compartment and both electrodes submerged. Avoid touching the bilayer itself, or walls of the insert, as this risks damaging the cells and generating inconsistent readings. This is even more important when using 12 mm diameter inserts, as variation is greater. If readings are inconsistent for a particular well, or the resistance is unexpectedly high, consider using fine grade sandpaper to carefully rub the silver tip of the electrodes (the bulbous, inward facing area of the chopsticks). No other part of the electrode should be touched – the chopsticks are incredibly delicate, so store carefully and avoid creasing the wire.\n\nNetTEER(Ω)=MeasuredTEER(Ω)−ReferenceTEER(Ω)(Equation 2)\n\nFinalTEER(Ω/cm2)=NetTEER(Ω)×AreaofTranswellinsert(cm2)(Equation 3)\n\nStep 6: Dextran Blue Studies (Day 14). Dextran blue permeability may be performed either at a single time point, as a measure of layer integrity, at multiple time points using single measurements, or at multiple time points sampling from the same well, enabling the calculation of ‘apparent permeability’. In the current studies, we have been comparing bilayers with single layers of both BPAEC and BATII, therefore we have chosen to use single sampling, single time point analysis. Unlike TEER, dextran blue studies mean sacrificing the insert used; therefore, this should be allowed for when calculating replicates and so on for bilayer studies.\n\nA 10 mg/mL stock solution should be prepared using Dextran Blue 2000 (DB2000; D5751, Sigma, MI, USA) in DPBS, alongside standard solutions of 7.5, 5.0, 3.75, 2.5, 1.25, 0.625, 0.3125 and 0.156 mg/mL. Add DPBS (600 μL) to each well of a fresh 24 well plate and equilibrate the plate to 37°C (minimum 15 minutes).\n\nRemove inserts containing cells one by one from the original plate, removing any apical medium during the transfer. Add the dextran blue stock solution (100 μL of 10 mg/mL) to the apical surface of each insert during the transfer and place the insert in the fresh plate. A blank insert is set up alongside the bilayer/monolayer cultures and used as a no cell control, along with a well containing 600 μL of DPBS and 100 μL dextran blue solution at 10 mg/mL, to determine extent of resistance exerted by the membrane.\n\nTechnical tip: Use two 200 μL pipettes when removing medium and adding dextran blue solution. This enables the user to have one pipette set to 100 μL (for dextran blue) and the other to 200 μL, to ensure that all apical medium is removed (if present – bilayers should have very little at 14 days).\n\nReturn the plate to 37°C, 5 % CO2 for 1 hour. At the end of the incubation, remove 600 μL from the basolateral compartment and read at 630 nm on a spectrophotometer in a 1 mL volume disposable cuvette. This should be carried out alongside 600 μL of the prepared standards. To allow for lower volumes in the apical compartments, we dilute an aliquot (60 μL) from apical solutions in 540 μL DPBS (a ten-fold dilution), before reading at 630 nm. Unknown values may then be extrapolated from a linear regression of the standard curve and corrected for dilutions accordingly.\n\nStep 7: IF Microscopy of whole inserts (Day 14 onwards). Cells cultured for 14 days are used for IF studies, to ensure that structural features reminiscent of the native alveolar epithelium are present, in particular a pseudostratified mix of ATII and their differentiated counterparts, ATI. At this stage, the endothelial layer of BPAECs have started to migrate through the membrane pores. This allows a more consistent contact with the underlying medium and also has the effect of providing ‘anchorage’. For this reason, we do not use an ECM to coat our Transwells, although we have included the protocol in our Methods, for users who wish to culture ATII cells or cell lines as a monolayer.\n\nTechnical tips: It is possible to divide membranes to maximise output; however, in our experience, the medial area of the insert is very delicate, so caution should be exercised and a fresh scalpel or razor blade used for each membrane, rocking the blade across the membrane rather than slicing, to reduce shear forces. When probing whole inserts, this may be done in situ, adding 100 μL of antibody solution to the apical side and 50 μL to the underside, holding the tip against the outer edge of the vertical wall of the insert and allowing the solution to be drawn under the membrane by capillary action.\n\nTranswell membranes are first washed three times (gently, holding the pipette against the inner wall of the insert when washing the apical surface) with DPBS. Cultures are then fixed in 4 % PFA for 15 – 30 minutes at room temperature. Each membrane is then washed three times in immunofluorescence buffer (IF buffer, details in Table 2), 5 minutes per wash. This also permeabilises the cells.\n\nPermeabilised cultures are blocked in DPBS/ 5 % normal goat serum (S26-M, Sigma Aldrich)/ 0.1 % triton X-100 (X100, Sigma Aldrich) for 1 hour, at room temperature. All antibody solutions are then made up in the same blocking buffer and applied to the apical and basolateral aspects of the membrane (see tips above). The plate containing the inserts are then incubated overnight at 4°C. The next day, inserts are washed three times with IF buffer (5 minutes per wash at room temperature) and secondary antibody applied for 1 hour at room temperature in the dark. Secondary antibodies are applied as follows: cells are again washed three times with IF buffer (5 minutes per wash at room temperature) and the membrane removed from the insert, using a fresh scalpel blade. Resting the inverted insert on a flat surface, run the leading edge of the blade around the circumference of the insert, turning the insert to complete the circuit, rather than moving the blade. This minimises ruffling of the membrane and serration of the membrane edge. The membrane is then placed onto a coverslip for mounting, apical side up. Place one drop of Prolong® Gold Antifade reagent with 4′, 6-diamidino-2-phenylindole (DAPI) (P36941, Thermo Fisher Scientific, Waltham, MA) or Vectashield Hardset mounting medium with phalloidin (TRITC) (H-1600, Vector Laboratories, Burlingame, CA, US) onto the centre of the membrane. Overlay the sample with a second clean coverslip, taking care not to introduce bubbles. Imaging may be performed after curing, using a confocal microscope. Primary antibodies are detailed in Table 4.\n\nNote: The Transwell insert membrane does auto-fluoresce in the near UV excitation spectrum, therefore may exert interference with DAPI staining, in the form of strong background signal. If this occurs, another nuclear stain should be used if required.\n\nStep 8: Fixation, embedding and haematoxylin & eosin (H&E) staining (Day 14 onwards). Inserts need to be carefully treated due to the delicate nature of bilayers. For this reason, all fixation, embedding and staining procedures are performed manually. Fix the inserts by immersion under 1 mL 10 % Formalin, divided between the apical and basolateral compartments. Avoid fixing for greater than 24 hours since tissue antigens may either be masked or destroyed. Thirty minutes is adequate for insert membranes. Because paraffin is immiscible with water, tissue must be dehydrated at room temperature, before adding molten paraffin wax. To do this, immerse in 70 % ethanol for 10 minutes, in 90 % ethanol for 10 minutes, in 100 % ethanol for 10 minutes and finally in isopropyl alcohol (IPA) 100 % for 5 minutes.\n\nTechnical tip: Using IPA instead of xylene allows the steps to be performed with the insert in situ, which makes handling and orientation easier.\n\nPlace the whole insert into IPA/paraffin 1:1 for up to 1 hour at 58°C. We perform this step at the embedding station, using pre-warmed IPA and pre-mixing IPA with paraffin in a vessel large enough to accommodate one insert, but small enough to sit in the molten wax compartment. After 1 hour, embed the tissue in paraffin at 58°C for another hour. At this stage, the membrane should be removed from the insert with a fresh scalpel, taking care to keep the insert apical side up in order to minimise damage to cells. Removed membranes are then embedded in fresh paraffin in a mould deep enough to accommodate them, with the thin edge perpendicular to the edge of the microtome blade. The block is allowed to cool at the embedding station, then stored at room temperature or 2–8°C until sectioning.\n\nFor best results, perform sectioning after blocks have been on ice (fill a suitable container with water and freeze at -20°C to form a solid block of ice) for 30 minutes. Cut 4–8 µm thick sections using a rotary microtome and float the sections onto a pre-heated 56°C water bath. Collect sections on histological slides (Superfrost +/+) and dry the slides overnight at 37°C. Slides can then be stored, either at room temperature or at 2–8°C. for several years in slide storage boxes.\n\nHaemotoxylin and Eosin (H&E) Staining. The staining procedure is outlined in detail in Lee et al.30, therefore this section details technical tips only.\n\nTechnical tip 1: All dewaxing and staining steps are carried out manually in a fume hood, since cells in a bilayer are prone to detaching during staining/wash procedures in automated systems.\n\nTechnical tip 2: Sections cut from embedded membranes are best cut at a minimum of 4 μm. We have also successfully stained 8 μm sections, which are less susceptible to detachment from the slide; however, this does affect clarity of images.\n\nTechnical tip 3: Rinsing slides under running tap water is best performed by filling a sandwich tub or similar with tap water and lowering a hose from the tap into the water, with the tap turned on to a trickle. Once confident that the water pressure is as low as possible, the user can then lower the cradle with slides into the tub. If this is still too aggressive and sections are detaching from the membrane or slide, we have successfully substituted running water for repeated rinsing in static tap water. To perform the rinsing step in this way, simply fill the tub with tap water and lower the cradle into the water without the hose. Leave for 5 minutes and gently lift the cradle out. Empty the tub and repeat the process until no stain is visible in the water (normally three to four washes).\n\n\nResults\n\nWe have isolated ATII cells as outlined in Lee et al.30 (Figure 1) and immortalised them using two gene combinations through lentiviral transduction. The use of the stem cell regulator and proto-oncogene Bmi1 in conjunction with hTERT was used by Fulcher et al. to generate a novel human bronchial epithelial cell line33 with considerable success. Like Fulcher et al., we found that the Bmi1/hTERT transduced B2AE cell line maintained genomic stability for up to at least 22 passages (Figure 2A). In addition, the proliferation rate was almost identical to that of the wild-type ATII cell (Figure 2B). Conversely, the SV40/hTERT transduced BATII cell line shows signs of genomic instability (Figure 2A) characteristic of viral oncogene derived cell lines34,35 at later passages, with a corresponding increase in proliferation rate by passage 10 (Figure 2B).\n\nFor karyotype analysis, a metaphase spread was performed on Colcemid cell cycle arrested cultures for both BATII and B2AE, comparing with the wild-type ATII (A). The Bmi1/hTERT transduced B2AE cell line was found to maintain the wild type bovine karyotype of 2n = 60, whilst the SV40/hTERT transduced BATII cell line showed signs of genomic instability at both passages 14 and 23. This was reflected in a growth curve analysis of the two cell lines, which was performed by seeding primary ATII cells into a 96 well plate, comparing over 9 days with the BATII and B2AE cell lines (B), using the Cell Titre Aqueous One Assay (Promega). Later passages (p10 and p23) of BATII cells exhibited a considerably higher rate of proliferation than the wild type ATII, whilst the B2AE cell line at all passages was comparable with its wild type counterpart. Data shown as optical density read at 490 nm (OD490nm); Mean ± SD, n=3.\n\nWe analysed monolayers of each cell line alongside the wild-type ATII, as cultures seeded onto Lab-Tek II coverglass. The phosphate transporter SLC34A2 is expressed exclusively in the ATII cell in lung tissue and is therefore considered a marker for ATII cells36–38. Both BATII and B2AE cells formed colonies (Figure 3A); however, the BATII cells in particular formed three-dimensional structures (Figure 3B, middle row). These were flatter on the periphery and raised in the centre. The outside edge of the BATII colonies exhibited a flattened morphology. This corresponded to a strong signal for the mesenchymal marker vimentin, the expression of which has been previously reported as being instrumental to the role of ATII progenitor cells in wound healing39 and indicative of the upregulation of mesenchymal markers in ATII cells cultured under two-dimensional conditons40. Whilst the ATII marker SLC34A2 was notably absent from the cells on the outside edge, it was very much evident in the adjacent cells of the colony, which formed the inner margins.\n\nLight microscopy images of the wild type ATII, SV40/hTERT lentivirus transduced BATII and Bmi1/hTERT lentivirus transduced B2AE were generated from cultures seeded into T75 cm2 flasks and imaged on a Nikon Eclipse Ci upright microscope (A). Scale bar 50 μm. For IF, each cell type was seeded onto a Nunc Lab Tek II coverglass, before fixation and staining with the ATII marker SLC434A2 and vimentin, a marker of epithelial-mesenchymal transition (B). Both cell lines were more prone to colony formation than their wild-type counterpart, with BATII cells in particular forming 3D structures, whilst B2AE frequently formed ring-like colonies around a central ‘lumen’. Cells which exhibited a flattened morphology on the exterior of the colonies of both cell lines stained most strongly for vimentin, with one lone cell in the BATII image notable for the complete absence of SLC34A2 (middle panel, white arrow), which remained strong in the centre of the 3D colony. Scale bar = 100 μm.\n\nThe concomitant increase of vimentin expression and decrease in SLC34A2 expression was most notable in the isolated colony of the BATII culture (Figure 3B, middle row, white arrow). This can be seen in more detail in the movie, presented in Dataset 1, see Extended data41. The B2AE colonies were flatter in nature (Figure 3B, bottom row) and thus more mesenchymal in terms of protein expression, with visibly reduced SLC34A2 expression throughout with vimentin expression that was comparable to the wild type ATII cultures. This observation, reflective of the flatter nature of the colonies in these cultures, indicates that the B2AE cell line differentiates readily to ATI, taking on the mesenchymal phenotype.\n\nPrevious literature has explored the role of extracellular matrix in the culture of epithelial cells of the distal airways40,42. We therefore explored the use of Matrigel in our early BATII cultures, coating 2D plastic and Transwell inserts with a 10 % solution of Matrigel in growth medium. Whilst cells seeded onto plastic (24 well format) formed the colonies thus far characteristic for this novel cell line (Figure 4A), the same cells, when seeded onto plastic coated with a 10 % solution of Matrigel, formed a sheet of epithelial cells (Figure 4B–C), which eventually peeled away from the sides of the well. These sheets contained 3D structures, a feature even more prominent when the cells were seeded onto 12 mm diameter 3 μm pore size Transwells (Figure 4D–H). The spheroidal, cyst-like structures seen in Figure 4D appeared to contain debris in the ‘lumen’ (Figure 4E). The structure was found to strongly express the ATII marker SLC34A2 around the periphery, whilst expression was notably absent in the ‘lumen’ (Figure 4F). When allowed to develop further (> 7 days), the sheet peeled back from the periphery of the Transwell insert membrane and formed a larger organoid structure visible with the naked eye (Figure 4G). This was embedded and sectioned for H&E staining and found to contain a live periphery (Figure 4H, black arrows) and a necrotic core (Figure 4H, ‘N’). Whilst these cultures exhibited an ATII phenotype, they were not suitable for transport experiments, since they could not be cultured at air-liquid interface as a result of the peeling away from the Transwell edge. Extracellular matrix (ECM) is not necessary for the culture of bilayers, since endothelial cells produce a laminin-rich ECM in culture43.\n\nBATII cells (p11) were grown on uncoated 2D plastic (A), 2D plastic coated in 10 % Matrigel (B–C) and a 12 mm diameter Transwell insert (D–H). When cultured on uncoated plastic, BATII cells formed colonies (A). On 10 % Matrigel coated plastic, the cells formed sheets (B), containing 3D spheroid structures (C). When cultured on Transwell inserts, these spheroids developed into organoid like structures (D–E). Immunofluorescence probing for the ATII marker SLC34A2 revealed that the organoid like structure was strong for this ATII marker at the periphery, whilst the ‘lumen’ was devoid of any expression (F). Embedded and sectioned organoids (G) were stained with haematoxylin and eosin (H), which revealed a central, necrotic area, N, with live cells at the periphery (arrows). Scale bars are 50μm (A), 50 μm (C), 100 μm (D), 25 μm (E–F) and 50 μm (H). The insert pictured in (C) is 12 mm diameter, with the organoid estimated at 3 mm diameter.\n\nBilayers were seeded for both novel ATII derived epithelial cell lines for comparison and cultured for 14 days, before probing for the ATII surface marker CD74 and mounting in TRITC-labelled phalloidin to highlight cell morphology and bilayer thickness by F-actin visualisation. The apical aspect of the BATII bilayer (Figure 5A) revealed 3D structures, containing cells which strongly expressed CD74 (white arrows), whilst the surrounding flatter areas only stained positive for phalloidin. The basolateral aspect (Figure 5B) also stained strongly for F-actin and demonstrated the difference in morphology between the ATII derived BATII cells and the endothelial BPAECs. These can be observed in more detail in Dataset 2, see Extended data41. Similar differences were observed in the phalloidin staining between the B2AE apical (Figure 5C) and basolateral (Figure 5D) aspects. The B2AE cell line did not appear to form 3D structures and CD74 expression appeared to be absent in these bilayers. Whilst B2AE appear to lose the ability to express CD74 when cultured at ALI, they do retain the ability to express the ATII marker surfactant protein C (Figure 5E). The endothelial surface marker CD31 was used to distinguish the BPAEC endothelial layer from the B2AE epithelial layer (Figure 5F); however, the poor signal and flattened morphology rendered it difficult to distinguish this cell line from the epithelial layer using endothelial specific markers.\n\nBilayers consisting of either the BATII (A) or B2AE (C) epithelial cell line overlaid onto a BPAEC (B and D) endothelial layer were cultured for 14 days at air-liquid interface (ALI) and stained for the ATII marker CD74 and phalloidin (A–D). The B2AE bilayer was also stained for the ATII cell marker Pro-SPC and the endothelial cell marker CD31 (E–F). BATII cells formed multiple 3D structures on the apical surface of the bilayer, the centre of which stained strongly for the ATII marker CD74 (A, arrow). Conversely, BPAEC cells formed a thin and flattened layer, which partially migrated through to the basolateral aspect of the insert membrane (B), the pores of which can be seen in the image and are thus used as point of reference when moving through the Z-stack. The B2AE cell line seemed to form a more uniform epithelial layer (C), which stained poorly for CD74. This did not affect the morphology of the BPAEC layer (D), which also migrated through the pores, to the basolateral aspect. Whilst B2AE cells do not stain strongly for CD74, they do retain the classical ATII marker, Pro-SPC (E), whilst CD31 was used to highlight the endothelial BPAEC layer (F). Below each image is the corresponding Z zoom (500 %) segment of the orthographic slice, including the Z slice indicator (yellow line). Scale bar = 100 μm.\n\nWe used haematoxylin and eosin (H&E) staining to study the morphology of the bilayers in 4 μm cross-sections taken from paraffin embedded Transwell inserts (Figure 6). Images clearly demonstrate the presence of two distinguishable layers on the apical and basolateral aspects of the membrane in both BATII (Figure 6A) and B2AE (Figure 6B) bilayers; however, detachment from the membrane during H&E processing was an issue and consequently, high-resolution images could not be obtained due to the distance between detached layers and the membrane. A repeat of the staining on sections acquired from bilayers seeded onto Transwell 12 mm diameter inserts with 0.4 μm pores enabled the visualisation of the cuboidal BATII cells (Figure 6C) as a distinct layer from the underlying endothelial BPAECs, whilst the B2AE exhibited a more flattened morphology (Figure 6D), characteristic of the ATII descendants, the alveolar type I (ATI).\n\nBilayers containing both cell lines were also stained for morphological analysis using haematoxylin and eosin. An apical and basolateral layer is clearly observed in both the BATII (A) and B2AE (B) bilayers (magnification 10 x). The fragility of the membrane during processing meant that it was difficult to visualise the intact bilayer and so a bilayer was seeded onto a Transwell 12 mm diameter insert with 0.4 μm pore size. This reduced the likelihood of cell detachment, as the endothelial cells remained on the apical aspect of the membrane, resulting in images where the epithelial BATII (C) and B2AE (D) layers are distinguishable from the underlying BPAEC, in which the staining is less intense (magnification 60 x). Cuboidal cells are visible particularly on the apical surface of the BATII bilayer (C, arrows), whilst the B2AE appear to be flatter and more reminiscent of the squamous alveolar type I (ATI) cells. Visualisation by light microscopy of the BATII/BPAEC bilayer shows the formation of the pseudostratified BATII epithelial layer on the apical surface (E); examples of raised areas denoted by arrows, in contrast to the underlying endothelial BPAEC monolayer (F) (magnification 20 x).\n\nThe integrity of bilayers containing BATII (A) and B2AE (B) epithelial cells overlaid onto BPAEC endothelial cells were measured by dextran blue 2000 (DB2000) permeability, following 14 days culture at air-liquid interface (ALI) on 6.5 mm diameter, 8 μm pore size Transwell membranes. Bilayers incorporating BATII cells as the epithelial layer significantly reduced apical to basolateral transport of DB2000 by almost 100 % (ANOVA, with Tukey’s Multiple Comparisons test; ** P ≤ 0.0001), whilst B2AE bilayers reduced permeability by 65 % (** P ≤ 0.0001). Whilst this was significant compared with inserts seeded with BPAEC cells only, this was a comparatively lower effect. Data shown is combined from three experiments (Mean ± SD; n= 9 inserts per group) performed on different days. The increased permeability of DB2000 through B2AE bilayers was reflected in the trans-epithelial electrical resistance (TEER) (C). Readings were significantly lower for B2AE bilayers from day 3 (unpaired t-test, P ≤ 0.0001). Combined data shown from three experiments performed on different days, displayed as Mean ± SD; n=18 (inserts per group).\n\nUsing light microscopy, it was possible to visualise the pseudostratified morphology developed by the BATII epithelial layer when cultured as part of a bilayer (Figure 6E). This was in contrast to the underlying BPAEC, which formed a flattened monolayer (Figure 6F).\n\nUsing the naked eye, it was possible to visualise the presence of any medium which had seeped through from the basolateral chamber to the apical side of the membrane. The BATII-containing bilayers were at air-liquid interface (ALI) from around 7 days, with less than 20 μL medium present in the apical chamber on feeding days. Further evidence of integral tight junction formation was generated in the significant reduction of dextran blue 2000 (DB2000) permeability across bilayers seeded onto 6.5 mm diameter Transwell inserts (Figure 7A and Dataset 3, see Underlying data41); this was an average (mean) of 96 % for the BATII derived bilayer (assigning blank insert as 0 % reduction), compared to 75 % reduction imparted by the BATII only monolayer (Tukey’s multiple comparisons test; P ≤ 0.0001) and 56 % for the endothelial BPAECs (P ≤ 0.0001). This was in stark contrast to the B2AE derived bilayer, which reduced DB2000 permeability by 65 %, compared to a 53 % reduction by the B2AE only monolayer (Figure 7B) (P ≤ 0.0001). There was no significant difference between the B2AE and BPAEC monolayers, the latter of which reduced DB2000 permeability by 48 %. This suggests B2AE cells formed a less integrated layer than BATII cells, supported by the considerably lower TEER values of B2AE derived bilayers, when compared to BATII derived bilayers, obtained over 21 days after seeding the epithelial layer (Figure 7C and Dataset 4, see Underlying data41). This finding was significant (unpaired t-test; P ≤ 0.0001).\n\n\nDiscussion\n\nWe describe here a method by which researchers may construct a bilayer for medium term culture, consisting of an immortalised alveolar type II (using the BATII novel cell line) epithelial cell layer overlaid onto a bovine pulmonary arterial endothelial cell (BPAEC) layer. Together, these two cell types recapitulate the highly gas permeable bilayer observed in vivo in the bovine alveolus. This is the first bovine bilayer model of the alveolus presented in the literature to our knowledge and thus has vast implications for the progression of research into bovine respiratory diseases, zoonoses and host pathogen interactions of the distal airways.\n\nIn order to reduce inter-experimental variability and therefore provide consistent data that are comparable with other studies, we opted to immortalise the ATII cells isolated according to our previously published protocol30. Conventionally, immortalisation involves the introduction of viral oncogenes, for example those encoding the E6 and E7 proteins of human papilloma virus (HPVE6/E7) and the large tumour (T) antigen derived from Simian virus (SV40). These, however, result in the trade off of genomic instability with the increased proliferation rate, which arises through the loss of the p53 checkpoint44, amongst other notable gene expression anomalies and an aberrant karyotype34,35,45. For this reason, we chose to generate an additional cell line, combining the proto-oncogene and stem cell regulator B cell-specific Moloney murine leukemia virus integration site 1, Bmi1 with the catalytic subunit of human telomerase, hTERT, to enable comparison with an SV40/hTERT immortalised cell line. The introduction of hTERT alone extends proliferative capacity; however, it does not fully immortalise cells46,47. Conversely, the combination of hTERT with Bmi1 has been previously reported to result in a genetically stable human bronchial epithelial cell (HBEC) line33, with the ability to form a differentiated, pseudostratified epithelium when grown at air-liquid interface (ALI), with a similar proliferation rate to non-immortalised cells. More recently, this combination was also used to successfully generate a human alveolar type II cell line48. Other alternatives to viral oncogenes include cyclin-dependant kinase 4 (Cdk4), as used by Ramirez et al.46. In their characterisation of HBECs immortalised using the combination of hTERT and Cdk4, Ramirez et al. do report regions of duplication (chromosomes 5 and 20). The gene expression profile, however, closely aligns with the wild-type and therefore this combination may yield another useful cell line derived from ATII cells.\n\nIn our studies, the combination of Bmi1 and hTERT in the generation of the B2AE cell line by lentiviral transduction did indeed yield a proliferation rate comparable to ATII cells and a stable karyotype. However, our intentions to use the cells in a medium-throughput assay to study host-pathogen interactions were hampered by the slow growth. This, together with the readiness of the B2AE to develop a squamous morphology indicative of differentiated alveolar type I (ATI) cells on plastic and when cultured as part of a bilayer made this cell line less suitable for studies of the role of ATII cells in innate immunity under ALI growth conditions.\n\nContinuous epithelial cell lines are often reduced in their differentiation capacity (for example the BEAS-2B line49), sometimes do not form the polarised layers characteristic of epithelial cells (for example the 16HBE140- line50) or demonstrate a reduced ability to form tight junctions (for example the A549 line)51. It was therefore important to characterise both the BATII and B2AE cell lines for these traits. The B2AE cell line showed a reduced propensity to develop tight junctions, as indicated by the formation of a relatively inefficient barrier to DB2000 when compared to the BATII cell line and relatively low TEER measurements over 21 days, when cultured as the epithelial component of a bilayer on Transwell inserts (Figure 7C and Dataset 4, see Underlying data41). A recently reported human alveolar epithelial cell line generated by the introduction of Bmi1 and hTERT48 reported TEER values of up to 400 Ω.cm2. Our own findings were that the B2AE cell line generated a maximum TEER of circa 500 Ω.cm2 by day 10 of culture as part of a bilayer. This was insufficient to maintain ALI beyond a couple of hours; therefore, the majority of culture time of B2AE bilayers was as a submerged culture. By contrast, the BATII cell line reached values in excess of 1200 Ω.cm2. These were comparable with rat primary alveolar epithelial cells and considerably higher than the human ATI-like A549 line52. BPAEC cells alone did not generate TEER values of greater than 200 Ω.cm2, suggesting they contribute relatively little to the impermeability of the bilayer.\n\nIt is possible that in our studies, the repression of p53 by the large T antigen of SV4044 has had a serendipitous beneficial effect. Certainly, when cultured at ALI in the presence or absence of BPAECs, the BATII cell line retains the ATII markers CD74 and SLC34A2 to a much greater degree than the B2AE cell line. Previous studies have reported that CD74 is lost quickly in the culture of primary ATII cells as they differentiate into ATI2,53. The BATII cell line appeared to retain expression of CD74 after two weeks culture at ALI, along with a pseudostratified morphology. It has been proposed that CD74 is a marker of proliferative capacity in ATII cells and that expression of this surface marker enables interaction of ATII cells with the macrophage migration inhibitory factor (MIF), stimulating ATII proliferation and alveolar repair in response to the loss of ATI cells2. Our findings support these data, since the B2AE cell line in our bilayer loses CD74 expression and takes on a squamous appearance, simultaneously failing to form a sufficiently integrated layer to allow culture at ALI. Taking these comparisons into account, we plan to focus our efforts on the BATII cell line for our own studies, which rely on a maintenance of the ATII phenotype by a sub-population of the cells when cultured as part of a bilayer. We do, however, propose that other researchers may find the B2AE cell line useful, such as in studies where a native karyotype is crucial.\n\nThe ability of the BATII cell line to maintain proliferative capacity and the expression of ATII markers, such as CD74 and SLC34A2, in 3D structures such as those observed in our Matrigel cultures alludes to their potential in spheroid and organoid studies of the alveolar epithelium. These spheroids were reminiscent of those reported by Lee et al.53. As such, they may be of interest to researchers in the field of developmental biology, for example in exploring the mechanisms of alveolar differentiation via the Wnt pathway, building on previous organoid culture approaches54. Previous literature reports the generation of organoids predominantly from sacrificed animals, as reviewed by Barkauskas et al.55. Disadvantages of this approach include the cost and logistics of animal housing (a particularly significant hurdle when studying larger animals such as bovines), the ethical considerations of such methods due to the invasive nature of procedures used to simulate disease status and the reported necessity for mesenchymal ‘support’ cells for the formation of ATII-derived organoids in 3D culture56. The BATII cell line derived spheroids require further characterisation, but we suggest that they offer much potential in the discipline of 3D organoid studies of the distal lung. Since our ongoing studies require the culture of cells at ALI as part of a bilayer, we have ceased to use Matrigel in our bilayer cultures, since the formation of organoids in this system causes the epithelial layer to peel back from the edges of the insert, causing ‘leakiness’ of the model.\n\nThe formation of an integral bilayer with BPAECs lends the BATII cell line to pharmacological studies57, or migration studies, including, but not exclusive to, neutrophils58. No such studies have been performed in a bovine bilayer thus far; therefore, we anticipate that the bilayer outlined here will facilitate bovine studies that parallel studies published previously for human in vitro models. The model enables users to obtain representative alveolar bilayers with desirable physiological and structural features of the native bovine alveolus within three weeks, whilst overcoming the challenges of housing large animals or the absence of large animal facilities. Most pertinent to the ethos of the 3Rs (in particular Replace), our bilayer model, ready to use in as little as three weeks, may be used as an in vitro alternative to in vivo experiments studying bovine respiratory diseases, for which 95 live cattle were used in the UK alone in 201629. These include, but are not exclusive to, bovine tuberculosis, bovine respiratory disease and bovine syncytial virus. Our method could also be used in parallel with live animal studies. For comparison with an in vivo scenario, one recent study in which cattle were infected with live M.tuberculosis was conducted using 12 calves reared from birth to six months of age in bio-containment level 3 laboratory facilities. Cattle were culled ten weeks post infection59, resulting in an overall study period of almost nine months. Whilst the number of live animals used may not be considered to be high, the authors would emphasise that the protracted nature of such experiments is a considerable drawback. It may also be argued that cattle used in such studies are not granted three of the ‘five freedoms of animal welfare’. These are “freedom from pain, injury or disease”, “freedom to express normal behaviour” and “freedom from fear and distress”25.\n\nOf particular interest to our group is the application of the model to study host-pathogen interactions in the bovine alveolus, between the ATII cell and M. bovis. Live animal studies have thus far provided a wealth of information, particularly those involving the natural target of M. bovis, the cattle themselves, as reviewed by Gregson et al.57. The close relationship between M. bovis and M. tuberculosis implies that an in vitro bovine bilayer model would be of use in elucidating data for both pathogens and therefore enable comparative studies in disease pathogenesis. This has certainly been demonstrated by the evaluation of human treatments and vaccines in the bovine live animal model22,23. In order to accelerate advances in therapies and control strategies, however, a more fundamental understanding of early disease pathogenesis is urgently required. Early events are difficult to study in live animal models, whereas the use of an in vitro bilayer could allow real time studies to be performed and would overcome the challenges of housing cattle in containment level 3 facilities.\n\nTo summarise, we present here two novel ATII derived cell lines, BATII and B2AE, together with a simple in vitro bilayer model of the bovine alveolus, including a detailed protocol by which other users can establish the model in their own institution. We have discussed the suitability of the immortalised BATII cells as organoid cultures to enable the translation of human developmental biology and tissue regeneration studies to the bovine model, whilst comparing and contrasting both cell lines generated by the lentiviral transduction of ATII cells with the viral oncogene SV40/hTERT combination (BATII), versus the proto-oncogene Bmi1/hTERT combination (B2AE). We have summarised how the use of the BATII cell line, in conjunction with the endothelial cell line BPAEC within a functional bilayer, could act as a viable alternative to live animal studies, overcoming logistical challenges associated with large animal models of disease pathogenesis. This applies particularly to pathogens such as M. bovis and M. tuberculosis, which require containment level 3 facilities. We propose that the bilayer model will allow the study of early host-pathogen interactions between the alveolar epithelium and the invading (respiratory) agent, including in the context of additional components of the innate immune system, such as neutrophils and macrophages. Such studies are extremely difficult to conceive in the in vivo setting due to the transient nature of such events. All of the components of the model are available either commercially (BPAECs, see ‘Culture conditions of BPAECs’) or through collaboration with the authors (BATII and B2AE cell lines). As such, other researchers are encouraged to consider this model as an alternative to live animal models. In doing so, there is considerable potential to advance the fields of respiratory disease in cattle, align research of bovine and human zoonoses and closely related pathogens in terms of preventative strategies as well as novel treatments and disease interventions. Consequently, the bilayer model offers great potential in replacing live animals used in the research of respiratory diseases affecting the distal airways. The authors are willing to discuss collaborations with other researchers wishing to use the bovine alveolar type II cell lines.\n\n\nData availability\n\nOpen Science Framework: f1000 research. https://doi.org/10.17605/OSF.IO/CXMBU41\n\nThe project contains the following underlying data:\n\n- Dataset 3_dextran blue assay.csv (Raw data underlying dextran blue permeability assay presented in Figure 7)\n\n- Dataset 4_TEER.csv (Raw TEER values underlying Figure 7)\n\nOpen Science Framework: f1000 research. https://doi.org/10.17605/OSF.IO/CXMBU41\n\nThe project contains the following extended data:\n\n- Dataset 1_3D BATII.mp4 (3D construction of the morphology of a BATII cell colony)\n\n- Dataset 2_Z stack bilayer.mp4 (Z stack of a bilayer composed of an apical BATII epithelial layer overlaid onto BPAECs)\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nThis study was supported by a strategic grant (NC/M002047/1), awarded by the National Centre for the Replacement, Refinement and Reduction of Animals in Research (NC3Rs).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nThe authors would like to thank members of the Veterinary Pathology Centre and Biochemical Sciences & Bioimaging Core Facility at the University of Surrey for their assistance.\n\n\nReferences\n\nVasilescu DM, Gao Z, Saha PK, et al.: Assessment of morphometry of pulmonary acini in mouse lungs by nondestructive imaging using multiscale microcomputed tomography. Proc Natl Acad Sci U S A. 2012; 109(42): 17105–10. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMarsh LM, Cakarova L, Kwapiszewska G, et al.: Surface expression of CD74 by type II alveolar epithelial cells: a potential mechanism for macrophage migration inhibitory factor-induced epithelial repair. Am J Physiol Lung Cell Mol Physiol. 2009; 296(3): L442–52. PubMed Abstract \| Publisher Full Text\n\nChen J, Chen Z, Narasaraju T, et al.: Isolation of highly pure alveolar epithelial type I and type II cells from rat lungs. Lab Invest. 2004; 84(6): 727–35. PubMed Abstract \| Publisher Full Text\n\nChen Z, Jin N, Narasaraju T, et al.: Identification of two novel markers for alveolar epithelial type I and II cells. Biochem Biophys Res Commun. 2004; 319(3): 774–80. PubMed Abstract \| Publisher Full Text\n\nJohnson MD, Widdicombe JH, Allen L, et al.: Alveolar epithelial type I cells contain transport proteins and transport sodium, supporting an active role for type I cells in regulation of lung liquid homeostasis. Proc Natl Acad Sci U S A. 2002; 99(4): 1966–71. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKapanci Y, Weibel ER, Kaplan HP, et al.: Pathogenesis and reversibility of the pulmonary lesions of oxygen toxicity in monkeys. II. Ultrastructural and morphometric studies. Lab Invest. 1969; 20(1): 101–18. PubMed Abstract\n\nKaplan HP, Robinson FR, Kapanci Y, et al.: Pathogenesis and reversibility of the pulmonary lesions of oxygen toxicity in monkeys. I. Clinical and light microscopic studies. Lab Invest. 1969; 20(1): 94–100. PubMed Abstract\n\nEvans MJ, Cabral LJ, Stephens RJ, et al.: Renewal of alveolar epithelium in the rat following exposure to NO2. Am J Pathol. 1973; 70(2): 175–98. PubMed Abstract \| Free Full Text\n\nBarkauskas CE, Cronce MJ, Rackley CR, et al.: Type 2 alveolar cells are stem cells in adult lung. J Clin Invest. 2013; 123(7): 3025–36. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFuchs S, Hollins AJ, Laue M, et al.: Differentiation of human alveolar epithelial cells in primary culture: morphological characterization and synthesis of caveolin-1 and surfactant protein-C. Cell Tissue Res. 2003; 311(1): 31–45. PubMed Abstract \| Publisher Full Text\n\nDegryse AL, Lawson WE: Progress toward improving animal models for idiopathic pulmonary fibrosis. Am J Med Sci. 2011; 341(6): 444–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMason RJ, Williams MC: Type II alveolar cell. Defender of the alveolus. Am Rev Respir Dis. 1977; 115(6 Pt 2): 81–91. PubMed Abstract\n\nWhitsett JA, Alenghat T: Respiratory epithelial cells orchestrate pulmonary innate immunity. Nat Immunol. 2015; 16(1): 27–35. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWhitsett JA, Wert SE, Weaver TE: Diseases of pulmonary surfactant homeostasis. Annu Rev Pathol. 2015; 10: 371–93. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNayak A, Dodagatta-Marri E, Tsolaki AG, et al.: An Insight into the Diverse Roles of Surfactant Proteins, SP-A and SP-D in Innate and Adaptive Immunity. Front Immunol. 2012; 3: 131. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCrouch E, Tu Y, Briner D, et al.: Ligand specificity of human surfactant protein D: expression of a mutant trimeric collectin that shows enhanced interactions with influenza A virus. J Biol Chem. 2005; 280(17): 17046–56. PubMed Abstract \| Publisher Full Text\n\nHartshorn KL, White MR, Tecle T, et al.: Innate defense against influenza A virus: activity of human neutrophil defensins and interactions of defensins with surfactant protein D. J Immunol. 2006; 176(11): 6962–72. PubMed Abstract \| Publisher Full Text\n\nOberley RE, Ault KA, Neff TL, et al.: Surfactant proteins A and D enhance the phagocytosis of Chlamydia into THP-1 cells. Am J Physiol Lung Cell Mol Physiol. 2004; 287(2): L296–306. PubMed Abstract \| Publisher Full Text\n\nKishore U, Greenhough TJ, Waters P, et al.: Surfactant proteins SP-A and SP-D: structure, function and receptors. Mol Immunol. 2006; 43(9): 1293–315. PubMed Abstract \| Publisher Full Text\n\nBirkness KA, Deslauriers M, Bartlett JH, et al.: An in vitro tissue culture bilayer model to examine early events in Mycobacterium tuberculosis infection. Infect Immun. 1999; 67(2): 653–8. PubMed Abstract \| Free Full Text\n\nWilliams A, Orme IM: Animal Models of Tuberculosis: An Overview. Microbiol Spectr. 2016; 4(4). PubMed Abstract \| Publisher Full Text\n\nDean GS, Rhodes SG, Coad M, et al.: Isoniazid treatment of Mycobacterium bovis in cattle as a model for human tuberculosis. Tuberculosis (Edinb). 2008; 88(6): 586–94. PubMed Abstract \| Publisher Full Text\n\nVordermeier HM, Villarreal-Ramos B, Cockle PJ, et al.: Viral booster vaccines improve Mycobacterium bovis BCG-induced protection against bovine tuberculosis. Infect Immun. 2009; 77(8): 3364–73. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWaters WR, Whelan AO, Lyashchenko KP, et al.: Immune responses in cattle inoculated with Mycobacterium bovis, Mycobacterium tuberculosis, or Mycobacterium kansasii. Clin Vaccine Immunol. 2010; 17(2): 247–52. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWebster J: Animal Welfare: Freedoms, Dominions and \"A Life Worth Living\". Animals (Basel). 2016; 6(6): pii: E35. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSu F, Liu X, Liu G, et al.: Establishment and evaluation of a stable cattle type II alveolar epithelial cell line. PLoS One. 2013; 8(9): e76036. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLee DF, Chambers MA: Isolation of Alveolar Type II Cells from Adult Bovine Lung. Curr Protoc Toxicol. 2019; e71. PubMed Abstract \| Publisher Full Text\n\nUrban-Chmiel R, Grooms DL: Prevention and control of bovine respiratory disease. J Livestock Sci. 2012; 3: 27–36. Reference Source\n\nOffice H: Annual Statistics of Scientific Procedures on Living Animals Great Britain 2016. Reference Source\n\nLee DF, Salguero FJ, Grainger D, et al.: Isolation and characterisation of alveolar type II pneumocytes from adult bovine lung. Sci Rep. 2018; 8(1): 11927. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYe J, Coulouris G, Zaretskaya I, et al.: Primer-BLAST: a tool to design target-specific primers for polymerase chain reaction. BMC Bioinformatics. 2012; 13: 134. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCampos PB, Sartore RC, Abdalla SN, et al.: Chromosomal spread preparation of human embryonic stem cells for karyotyping. J Vis Exp. 2009; (31): pii: 1512. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFulcher ML, Gabriel SE, Olsen JC, et al.: Novel human bronchial epithelial cell lines for cystic fibrosis research. Am J Physiol Lung Cell Mol Physiol. 2009; 296(1): L82–91. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWoods C, LeFeuvre C, Stewart N, et al.: Induction of genomic instability in SV40 transformed human cells: sufficiency of the N-terminal 147 amino acids of large T antigen and role of pRB and p53. Oncogene. 1994; 9(10): 2943–50. PubMed Abstract\n\nHein J, Boichuk S, Wu J, et al.: Simian virus 40 large T antigen disrupts genome integrity and activates a DNA damage response via Bub1 binding. J Virol. 2009; 83(1): 117–27. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYin BW, Kiyamova R, Chua R, et al.: Monoclonal antibody MX35 detects the membrane transporter NaPi2b (SLC34A2) in human carcinomas. Cancer Immun. 2008; 8: 3. PubMed Abstract \| Free Full Text\n\nHashimoto M, Wang DY, Kamo T, et al.: Isolation and localization of type IIb Na/Pi cotransporter in the developing rat lung. Am J Pathol. 2000; 157(1): 21–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTraebert M, Hattenhauer O, Murer H, et al.: Expression of type II Na-Pi cotransporter in alveolar type II cells. Am J Physiol. 1999; 277(5): L868–73. PubMed Abstract \| Publisher Full Text\n\nFujino N, Kubo H, Suzuki T, et al.: Isolation of alveolar epithelial type II progenitor cells from adult human lungs. Lab Invest. 2011; 91(3): 363–78. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCalle EA, Mendez JJ, Ghaedi M, et al.: Fate of distal lung epithelium cultured in a decellularized lung extracellular matrix. Tissue Eng Part A. 2015; 21(11–12): 1916–28. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLee D, Chambers M: f1000research. 2019. http://www.doi.org/10.17605/OSF.IO/CXMBU\n\nBhowmick R, Gappa-Fahlenkamp H: Cells and Culture Systems Used to Model the Small Airway Epithelium. Lung. 2016; 194(3): 419–28. PubMed Abstract \| Publisher Full Text\n\nDavis GE, Senger DR: Endothelial extracellular matrix: biosynthesis, remodeling, and functions during vascular morphogenesis and neovessel stabilization. Circ Res. 2005; 97(11): 1093–107. PubMed Abstract \| Publisher Full Text\n\nSheppard HM, Corneillie SI, Espiritu C, et al.: New insights into the mechanism of inhibition of p53 by simian virus 40 large T antigen. Mol Cell Biol. 1999; 19(4): 2746–53. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCudré-Mauroux C, Occhiodoro T, König S, et al.: Lentivector-mediated transfer of Bmi-1 and telomerase in muscle satellite cells yields a duchenne myoblast cell line with long-term genotypic and phenotypic stability. Hum Gene Ther. 2003; 14(16): 1525–33. PubMed Abstract \| Publisher Full Text\n\nRamirez RD, Sheridan S, Girard L, et al.: Immortalization of human bronchial epithelial cells in the absence of viral oncoproteins. Cancer Res. 2004; 64(24): 9027–34. PubMed Abstract \| Publisher Full Text\n\nHoutmeyers E, Gosselink R, Gayan-Ramirez G, et al.: Effects of drugs on mucus clearance. Eur Respir J. 1999; 14(2): 452–67. PubMed Abstract \| Publisher Full Text\n\nTamo L, Hibaoui Y, Kallol S, et al.: Generation of an alveolar epithelial type II cell line from induced pluripotent stem cells. Am J Physiol Lung Cell Mol Physiol. 2018; 315(6): L921–L932. PubMed Abstract \| Publisher Full Text\n\nStewart CE, Torr EE, Mohd Jamili NH, et al.: Evaluation of differentiated human bronchial epithelial cell culture systems for asthma research. J Allergy (Cairo). 2012; 2012: 943982. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nForbes B, Ehrhardt C: Human respiratory epithelial cell culture for drug delivery applications. Eur J Pharm Biopharm. 2005; 60(2): 193–205. PubMed Abstract \| Publisher Full Text\n\nCooney DJ, Hickey AJ: Cellular response to the deposition of diesel exhaust particle aerosols onto human lung cells grown at the air-liquid interface by inertial impaction. Toxicol In Vitro. 2011; 25(8): 1953–65. PubMed Abstract \| Publisher Full Text\n\nSrinivasan B, Kolli AR, Esch MB, et al.: TEER measurement techniques for in vitro barrier model systems. J Lab Autom. 2015; 20(2): 107–26. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLee JH, Kim J, Gludish D, et al.: Surfactant protein-C chromatin-bound green fluorescence protein reporter mice reveal heterogeneity of surfactant protein C-expressing lung cells. Am J Respir Cell Mol Biol. 2013; 48(3): 288–98. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLee JH, Tammela T, Hofree M, et al.: Anatomically and Functionally Distinct Lung Mesenchymal Populations Marked by Lgr5 and Lgr6. Cell. 2017; 170(6): 1149–1163.e12. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBarkauskas CE, Chung MI, Fioret B, et al.: Lung organoids: current uses and future promise. Development. 2017; 144(6): 986–997. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcQualter JL, Yuen K, Williams B, et al.: Evidence of an epithelial stem/progenitor cell hierarchy in the adult mouse lung. Proc Natl Acad Sci U S A. 2010; 107(4): 1414–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGregson L, Hope WW, Howard SJ: In vitro model of invasive pulmonary aspergillosis in the human alveolus. Methods Mol Biol. 2012; 845: 361–7. PubMed Abstract \| Publisher Full Text\n\nWeppler A, Rowter D, Hermanns I, et al.: Modulation of endotoxin-induced neutrophil transendothelial migration by alveolar epithelium in a defined bilayer model. Exp Lung Res. 2006; 32(10): 455–82. PubMed Abstract \| Publisher Full Text\n\nVillarreal-Ramos B, Berg S, Whelan A, et al.: Experimental infection of cattle with Mycobacterium tuberculosis isolates shows the attenuation of the human tubercle bacillus for cattle. Sci Rep. 2018; 8(1): 894. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "46557", "date": "05 Apr 2019", "name": "Pietro Mastroeni", "expertise": [ "Reviewer Expertise Infection", "Immunity", "Histopathology", "tissue microbiology", "molecular biology", "vaccines", "antibiotics. Referee suggested by the NC3Rs for their scientific expertise and experience in assessing 3Rs impact" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper describes the generation and characterisation of two immortalised bovine epithelial cell lines to be used in conjunction with bovine arterial endothelial cells for the creation of structures that would model the alveolar environment of bovine species. The authors very clearly list the potential applications of such approach and its impact on the 3Rs. It would be desirable to explain in more depth what is the incontrovertible evidence that such models are truly representative of an in vivo situation.\nThe introduction is very clear and comprehensive and allows the reader to gain information on the background to the study and on the general context of the experiments. The methods section is also clear and detailed, thus allowing reproducibility of the experiments.\n\nThere are a few points that the authors might want to address in their revision of the manuscript.\nFigure 3. The pictures relative to the wild type cells are hugely different from the ones obtained when using either of the cell lines. This might be confusing for the reader who may wonder how such large morphological differences can then lead to the conclusion that the immortalised cells lines are a good model for the natural alveolar environment. Perhaps this needs to be explained more clearly and discussed in detail. In some sections the authors describe results for only one cell line (e.g. Matrigel cultures) whilst in others they cover both cell lines. This may be confusing. A clear explanation of the choice of the cell markers is not always provided in the results section. This would make the manuscript clearer. It might be helpful to have a conclusive statement at the end of each section of the results to state the main findings and conclusions of that particular set of experiments. Similarly, a sentence introducing each section and explaining why the experiments were done would add clarity. This is done for some results sections but not for others. Therefore, sometimes the reader is left to wonder....why are they doing this? It might be useful to show some histological and immune-fluorescence pictures of alveolar structures from a bovine lung to show the extent of similarities and differences between the proposed approach and the mammalian alveolar structure. Alveolar macrophages (together with other immune cells) are a key to lung homeostasis and protection against infectious diseases. It needs explaining how structures devoid of such important immune cells can be considered representative of an in vivo situation, especially when one of the intended uses would be the study of infectious diseases. The approach described in this study entails the generation of cell lines from a single animal. How can the authors be sure that these cell lines would be representative especially when dealing with a species that is clearly outbred. What are the dangers of animal-to-animal variability and therefore lack of representativeness?\nAre a suitable application and appropriate end-users identified? Yes\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\nAre the 3Rs implications of the work described accurately? Yes\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4759", "date": "30 Jul 2019", "name": "Diane Lee", "role": "Author Response", "response": "The authors acknowledge that here, as in many cell lines, there are morphological differences when compared to the wild type when grown on coverglass. We attributed this to the increased proliferation rate of the BATII cell line and the lack of contact inhibition, characteristic of SV40 Large T Antigen transformed cells. The B2AE cell line was slower by comparison and did not form multi-layered structures on coverglass. Although we still observed multilayer structures in co-cultures generated using the BATII cell line (and to a lesser extent, the wild-type ATII), these covered <20 % of the culture surface, being surrounded by thinner mono/bilayers of epithelial-like cells. We have expanded the results section to clarify and also included a discussion of the advantages versus disadvantages of using cell lines which differ in phenotype from their wild-type counterparts. The omission of the B2AE cell line from some experiments was due to the growth rate of B2AE being very slow and not generating enough stocks to perform experiments with meaningful numbers of replicates. We chose to therefore focus on the BATII cell line for some experiments and have clarified as such in the text. We have added explanations or further clarification to justify choice of marker in each instance, along with references where necessary. Conclusive statements have also now been included or extended for the results sections, along with the inclusions of an introductory sentence or two. We have included an H&E image and highlighted differences/similarities between the proposed model and alveolar architecture in discussion. In addition, the inclusion of immune cells has been added to the discussion – studies including peripheral blood mononuclear cells (PBMCs) have been performed and are being published separately as part of further studies of M.bovis-BCG interaction with the bovine alveolus (manuscript in preparation). Lastly, the authors concede that this is from a single animal and therefore cannot be entirely representative of a whole species or even breed. We have added further to the discussion to this effect, with additional emphasis on the positives and negatives of the utilisation of cell lines when modelling the in vivo situation. We have also published detailed protocols of the isolation procedure, enabling researchers to isolate cells from multiple animals and compare phenotypes." } ] }, { "id": "46560", "date": "07 May 2019", "name": "Claus-Michael Lehr", "expertise": [ "Reviewer Expertise Advanced Drug Delivery Technologies", "cell and tissue models of epithelial barriers (intestines", "skin", "lung)" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors provide a detailed guide to reproduce an in vitro tissue model of the bovine air-blood barrier, that consists out of bovine pulmonary arterial endothelial cells in co-culture with one out of two different immortalized bovine alveolar epithelial cell lines. The introduction puts the experiments in a concise but at the same well explained scientific context. The methods section is easy to follow and should allow for the reproduction of the presented model, although it could profit from some revisions:\nThe diameter (mm) of all used Transwell growth supports should be exchanged by the growth area (cm²) to easily identify the growth area. Measurement of cell layer integrity: It is unclear if pre-warmed medium was used to restore submerged conditions and then TEER was measured after 15 min of equilibration. In our case submerged conditions are restored by adding 200 µl apical and 600 µl basolateral (in case of 0.33 cm² inserts) of pre-warmed (37°C) medium to the ALI cultures. TEER values are then measured after at least 1 h of equilibration. The TEER measurement is performed on a heating plate (37°C), to prevent a fast temperature drop.\n\nAlthough a detailed overview of possible applications for this model to aid the 3R’s in future research is given be authors, the conclusions drawn from the displayed often appear not to be justified based on the data presented. This especially accounts for the results as well as the discussion section. The scientific benefits claimed in the “Research Highlights” appear to be exaggerated and are not based on the actually provided data.\nResults section:\nIn accordance with the comments of Pietro Mastroeni, a scientific discussion of the selected markers is missing. Inspiration for discussion could be taken from references 11 and 22. Figure 6 A and B clearly show different magnifications Figure 6 C and D show the formation of multilayered epithelia and are not further discussed. BPAEC in Figure 7 A looks as the same as in Figure 7 B. Is it the same? Maybe combine into one figure.\nDiscussion section:\nThe model is termed a “bilayer” of epithelial and endothelial cells. As shown in Figure 6 C and D, the BAT II culture shows the formation of multiple stacked layers, the B2AE culture at least more than two layers. This issue is not properly mentioned in the discussion. The development of TEER values greater than 1000 Ohm*cm² for the BATII cells, could be explained by the increased resistance of paracellular flux based on the formation of functional tight junctions within a monolayer (as expected in vivo) or by an increase in resistance through the formation of multilayers. The formation of functional tight junctional complexes was in this study only indicated via TEER measurement and reduction of dextran permeability; immunohistochemistry or transcriptomic data are missing. Given the context of the massive multilayer formation shown in Fig. 6 C for the BATII culture, the missing information on tight junctional integrity would at least need a thorough discussion. The authors appear to be completely unaware of some quite relevant scientific work by other groups in the context of primary rat or human alveolar epithelial cells, resp. (See references 33 and 44) In the context of human alveolar epithelial cell lines the work of Tetley et al.5 Kuehn et al.6 should be considered. The same is true for the work as regards co-cultures of alveolar epithelial cells and immune cells by Rothen-Rutishauser et al.7 Hittinger et al.8 Kletting et. al.9. A very recent paper by Costa et al.10 describes a co-culture of human alveolar epithelial cells, endothelial cells and macrophages. The whole discussion could benefit of an even more critical reflection of the advantages as well as disadvantages of the proposed model.\nAre a suitable application and appropriate end-users identified? Yes\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\nAre the 3Rs implications of the work described accurately? Yes\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Partly\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly", "responses": [ { "c_id": "4760", "date": "30 Jul 2019", "name": "Diane Lee", "role": "Author Response", "response": "The authors have modified the manuscript to include the surface area of the Transwell inserts, in addition to the diameter throughout the manuscript. In performing TEER measurements, medium was pre-warmed (in the incubator). We do not have a heating plate and therefore some loss of temperature was inevitable. To help counteract this, we take three measurements from all wells (for each measurement working from the first to last sample before starting again at the first), to monitor the extent of TEER change with temperature. This has been clarified in the methods. We did calculate drift for eight inserts between the first and last measurements. In these calculations, TEER increased between 3 and 25 % (Mean 9 ± 8 %) for the BATII co-cultures and 3 and 22 % (Mean of 7 ± 6 %) for B2AE co-cultures respectively. Drift was highest in the earlier measurements (day 3) in both cases. Until co-culture studies using the proposed model are performed at length and data from multiple groups compared, it’s not possible to say for sure how much of a benefit any model will provide to the field. The authors continue to refine the model and encourage others to collaborate or make use of this approach in their own studies. With regards to scientific benefits, there are studies which simply cannot be performed in live bovines at present; these particularly include the early (and transient) events which occur during infection with pathogens such as M.bovis, or RSV. We hope that the benefits of such an approach as the one proposed in the current paper are clarified by our extended discussion. We have further clarified why we chose the markers used in this study and included further discussion as appropriate. We thank the observant reviewer for highlighting the discrepancy in Figure 6 and have corrected the figure accordingly. With regards to the comments on Figure 6 C and D, we have added to both the results section and discussion section with regards to the multi-layered epithelial layers in the bilayers generated by BATII. These were isolated in the ALI cultures, their presence being easily identified using x-z confocal microscopy, as shown in Figure 5 and the extended data. We have discussed these morphological anomalies further in the text. The data for BPAEC in Figure 7A and 7B are from different experiments/sets of data. The actual values are 55.7 ± 5.2 and 48 ± 3.4 % respectively. The authors concede that in areas, the model is more than two cells thick, although this is not universal across the membrane (shown by the x-z confocal images of Figure 5 and extended data). We have included a further image (E) in Figure 6 to put these 3D structures in context and expanded the discussion accordingly, including whether it would be more accurate to describe the model as ‘co-culture’. It should be noted, however, that in our studies, the wild type ATII cells also formed 3D structures when grown on membranes in co-culture with BPAEC (see Dataset 6, which shows frames from an ATII/BPAEC co-culture Z-stack). This occurred whether in co-culture with the endothelial BPAEC, as in the current study, or when seeded as a mono-culture, as in our previous paper (see Lee et al. 2018). We accept that the use of the term ‘bilayer’ in this instance is not accurate and have modified the manuscript to describe the model as a co-culture, rather than a bilayer. TEER values of >1000 were also generated for wild-type ATII co-cultures with endothelial cells, which we have included in the extended data (Dataset 5 and 7). As clarified above (and further clarified in the text), the 3D structures were not uniform in thickness. The thickness ranged from one cell upwards. It is our experience that TEER values represent the area of the lowest integrity in the culture i.e. the thinnest/with the fewest number of functional tight junctions. This is evidenced by the drop in TEER in cultures damaged by experimental manipulations (scrape by pipette and so on). Dataset 5 contains raw includes raw TEER data for BATII/B2AE only, BPAEC only, WT only and each as part of a co-culture of epithelial/endothelial cells, plotted as dataset 7. We have not included IHC data as the antibodies we tried (ZO-1 and Occludens) were cross-reactive with other antigens or generated high background signal. This lowered our confidence in their suitability for their use in the characterisation of the model. The authors are indeed aware of the work of Kemp et al. (referred to as Tetley above). Whilst they do consider the work relevant to discussions of immortalised cell lines, Kemp et al. transduced a temperature sensitive mutant of SV40 and cultured the resulting cells without other cell types. We have extended the discussion to further consider this and the work of other groups. We have recently completed studies of the interaction of the co-culture model of the current study with immune cells (PBMCs). As the current paper is a methods paper, we focussed here on the setup of the model and a manuscript detailing our latter work is in preparation." } ] }, { "id": "46561", "date": "31 May 2019", "name": "Robert L Davies", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper describes the establishment of a co-culture bilayer model of the bovine alveolus using bovine pulmonary arterial endothelial cells (BPAECs) and immortalised bovine alveolar type II epithelial cells cultured for 14 days at an air-liquid interface (ALI). The authors highlight the practical, scientific and 3Rs benefits of their model and provide examples of current and potential applications.\nThe rationale for establishing a bilayer alveolus model is well justified and the methodology is generally detailed and clear. The authors have chosen to create their bilayer by seeding the alveolar epithelial cells directly on top of confluent BPAECs and culturing at ALI (it is not clear whether an initial submerged phase was used which is normal practice). However, in previously established bilayer models of the alveolus (Hope et al., 20071; Morton et al., 20142; Costa et al., 20193), the endothelial and epithelial cells have been grown on the basal and apical surfaces of the membrane, respectively. This better mimics the uninterrupted interaction between the endothelial cells and tissue fluids (i.e. blood). It is not clear why the authors have taken the approach they have, but I think this is an important point that should be highlighted with reference to other published studies and discussed. Was there a specific reason this approach was chosen? In relation to this, in the 5-day and 14-day cultures shown in Figure 1, cells are shown on the basal surface of the membrane; this is confusing. If these represent BPAECs that have migrated to the underside of the membrane, this needs to be explained because, in effect, a double endothelial layer has been created which is not representative of the in vivo situation. Is this related to the use of 8 mm pores (which seems very large) and could this have been avoided by the use of a smaller pore size (which were used in other experiments - see also comment below)?\nThe reasoning behind the development of immortalized cell lines is logical although there are also disadvantages to this approach as demonstrated and discussed in the paper (the B2AE cell line, in particular, was not very successful). The imaging shown in Figure 3 highlights significant differences between the wild-type and immortalized cell lines grown on “coverglass”.\n\nUnfortunately, wild-type primary alveolar epithelial cells do not appear to have been used in subsequent bilayer experiments which would have allowed important direct comparisons to have been made between these and the immortalized cell lines. In particular, the histological section of the BATII/BPAEC layer shown in Figure 6C shows a thickened structure (and is certainly not pseudostratified) that is not representative of the native alveolar epithelium. The latter is obviously a very thin monolayer to allow for gaseous diffusion and, on this point, it would be helpful to have a figure of a histological section of ex vivo lung tissue showing the alveolar epithelium. Indeed, a more realistic single-cell layer is seen in the alveolar model created by Costa et al. (2019)3. It would be helpful to readers and potential users of the cell lines for these points to be discussed. In relation to these observations, the authors have switched between using membranes of different pore-diameters (0.4, 3.0 and 8 mm). Pore diameter is extremely important and selection is based on the experimental objectives. It is not clear why different pore size have been used and this could also be discussed.\n\nAre a suitable application and appropriate end-users identified? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre the 3Rs implications of the work described accurately? Yes\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4761", "date": "30 Jul 2019", "name": "Diane Lee", "role": "Author Response", "response": "Initial experiments did involve seeding cells on opposing sides of the membrane. Owing to difficulties generating quality sections of paraffin embedded membranes, we experimented with seeding both cell types on the apical side, as described by Birkness et al.(1999) and following personal communication with the corresponding author of that paper. Following this method allowed us to compare data, since our intention was to study the interaction of bovine TB with the alveolus, making our model the bovine equivalent of that of Birkness et al. On the point of mimicking the uninterrupted interaction between epithelial cells and endothelial cells, the authors suggest that in no in vivo situation will you find a synthetic membrane and that less interruption occurs when the cells are adjacent, rather than reaching through a pore. The large pores were necessary due to the intended use of the model, which would require mycobacteria and immune cells to migrate from one chamber to another. These points have now been discussed in the manuscript. The authors whole-heartedly agree that there are disadvantages as well as advantages to the use of immortalised cell lines. These have been included at length in the modified discussion. In light of the reviewer’s suggestions, we have included extended data with regards to the wild-type ATII as part of co-culture with endothelial cells (Dataset 6 shows IF images of z-stack slices, whilst dataset 7 includes raw data of the TEER values generated by culture and co-culture of each cell line and the wild-type ATII). The work of Costa et al. was published after the submission of version 1 of this manuscript, but has now been included in the discussion accordingly, as we agree that this is important to consider. We have also modified the manuscript discussion text to include justification for the different pore sizes. In brief, 0.4 was used to generate quality sections – a particular technical challenge, whilst 3 and 8 were used due to intended studies which included migration of PBMCs and mycobacteria (we notice that Costa et al. also use 3.0 μm, presumably for the same reason)." } ] } ]	1	https://f1000research.com/articles/8-357
https://f1000research.com/articles/8-1218/v1	30 Jul 19	{ "type": "Case Report", "title": "Case Report: rare hybrid lesion of a central giant cell granuloma within a juvenile ossifying fibroma", "authors": [ "Hadeer Rizk Saad", "Noura M. Kamal", "Hatem W. Amer", "Noura M. Kamal", "Hatem W. Amer" ], "abstract": "Central giant cell granuloma (CGCG) is classified by the World Health Organization as a benign bone lesion. It is found anteriorly in the mandible, with most of the cases crossing the midline. In total, 70% of CGCGs are encountered in young females. Fibro-osseous lesions are a group of pathologies that encompass neoplastic, dysplastic and reactive entities. Juvenile ossifying fibroma, which can be further categorized into juvenile trabecular ossifying fibroma (JTOF) and juvenile psammomatoid ossifying fibroma, represents an aggressive neoplastic example of these fibro-osseous lesions. JTOF occurs in children at almost equal ratios in both sexes, affecting the maxilla more than mandible. This study aims to report a peculiar case of a hybrid lesion comprising CGCG and JTOF in the mandible of a nine-year-old female patient. Clinical, radiographic and histopathological findings were assessed. Clinical examination revealed an intraoral swelling extending from the right impacted third molar area to the left first molar area. Computed tomography showed a well-defined multilocular radiolucency with diffuse flecks of radioopacities. Histopathologically, the lesion comprised fibrous connective tissue encompassing numerous multinucleated giant cells and other areas of cell-rich connective tissue stroma containing bands of osteoid matrix and anastomosing immature bone trabeculae intermixed with scattered clusters of multinucleated giant cells. We hereby report a case of a rare hybrid lesion comprising CGCG and JTOF.", "keywords": [ "Aggressive lesion", "Central giant cell granuloma", "Fibrous dysplasia", "Hybrid lesion", "Juvenile ossifying fibroma", "Pathology" ], "content": "Introduction\n\nCentral giant cell granuloma (CGCG) is believed to be an entity exclusive to jaw bones1. It occurs anteriorly in either jaw but is more commonly encountered in the mandible. The majority of cases manifest in children or young adults and are twice as frequent in females than males2,3. It attains a quite benign behavior, although rare aggressive lesions have been documented3.\n\nThe term “benign fibro-osseous lesions” generically describes an aggregate of pathologies that are characterized by replacement of normal bone with a highly cellular connective tissue associated with newly formed bone trabeculae4. These lesions encompass varying entities including neoplastic, dysplastic and reactive pathologic processes1. Ossifying fibroma (OF) and its subtypes represent their neoplastic entity. Juvenile OF (JOF) is an aggressive subtype of ossifying fibromas that can be further classified into juvenile trabecular OF (JTOF) and juvenile psammomatoid OF5. JTOF occurs in children and adolescents with equal distribution of cases among both sexes. It affects the maxilla more than mandible, while other skeletal bones are rare to be affected with such lesion3.\n\nRadiographically, CGCGs usually present as well-demarcated radiolucencies that are multilocular, although unilocular lesions have been noted6. As for JOFs, despite their non-encapsulated nature, they appear well delineated, with early lesions found as unilocular radiolucencies. However, old lesions more commonly assume the multilocular configuration7. Depending on the degree of ossification within the lesion, JOF can present varying extents of radioopacities within its radiographic picture. We can then clarify that those precedent features often challenge the diagnosis and management of such lesions. We hereby report a case of a hybrid lesion that contained both of the previously mentioned entities.\n\n\nCase report\n\nA nine-year-old female patient was reported to the Oral and Maxillofacial Surgery Department at the Faculty of Dentistry, Cairo University, complaining of a painless slowly growing swelling in the anterior area of the mandible of one year’s duration, causing buccal and lingual bone expansion, displacement and looseness of the incisors (Figure 1). Her medical and family histories were unremarkable. An incisional biopsy was performed, and gross examination revealed two firm pieces 2.5×2×1.2 cm in size, which were reddish white in color and solid in consistency (Figure 2).\n\nPreoperative surgical plan showing the swelling in the anterior area of the mandible, causing buccal and lingual bone expansion and displacement of the incisors.\n\nGross examination of the incisional biopsy revealed two firm pieces of 2.5×2×1.2 cm in size; they were reddish white in color and solid in consistency.\n\nThe histopathological examination using routine H&E staining demonstrated heavily scattered multinucleated giant cells in a background of highly cellular fibrous stroma consisting of mononuclear stromal cells and extravasated red blood cells. Newly formed bone trabeculae and osteoid tissue were noted at the periphery of the lesion (Figure 3). Accordingly, it was diagnosed as central giant cell granuloma. The surgeon consequently took the decision to inject the lesion with the corticosteroid triamcinolone acetonide (10 mg/ml) dissolved in lidocaine anesthetic solution in equal parts. It was administered in a dosage of 1 ml for every 1 cm3 of the lesion through 6 weekly injections, upon which the lesion showed marked regression.\n\n(a) Photomicrograph of old incisional biopsy showing heavily scattered multinucleated giant cells (short arrow) in a background of highly cellular fibrous stroma consisting of mononuclear stromal cells (long arrow) and extravasated red blood cells (original magnification 40×). (b) Newly formed bone trabeculae and osteoid tissue (star) were noted at the periphery of the lesion (original magnification 40×)\n\nSix months later, the swelling significantly increased in size, causing severe facial disfigurement. Intra-orally, it developed a brownish red discoloration and was non-tender but displaced the otherwise clinically sound related teeth. Computed tomography (CT) of the lesion showed it to be a well-defined multilocular radiolucency with diffuse flecks of radioopacities. It extended from the right impacted third molar area to the left first molar area causing perforations in the buccal and lingual cortical plates (Figure 4).\n\n(a) CT of the lesion showing a well-defined multilocular radiolucency with diffuse flecks of radiopacities, extending from the right impacted third molar area to the left first molar area. (b) CT of the lesion showing perforations in the buccal and lingual cortical plates.\n\nThe lesion underwent an excisional biopsy. On gross examination, it consisted of almost 30 pieces that ranged in size, with an average size of 2×2×1.5 cm. They were reddish in color, some were hard and others were soft in consistency (Figure 5). The specimen also contained multiple tooth follicles of both the primary and permanent dentitions. Sections of this biopsy revealed a highly cellular fibrous connective tissue encompassing numerous multinucleated giant cells. Areas of intralesional hemorrhage were also evident; two findings that are consistent with the diagnosis of central giant cell granuloma. However, other sections revealed extensive areas of cell-rich connective tissue stroma containing bands of osteoid matrix and anastomosing immature bone trabeculae surrounded by plump osteoblasts intermixed with scattered clusters of multinucleated giant cells (Figure 6), features that are characteristic of juvenile trabecular ossifying fibroma. We subsequently reached the final diagnosis of a hybrid lesion demonstrating central giant cell granuloma in occurrence with juvenile trabecular ossifying fibroma.\n\nGross examination of the excised lesion showing numerous, reddish, hard and soft pieces of tissue that ranged in size, having the average of 2×2×1.5 cm.\n\n(a) Photomicrograph of the excisional biopsy revealed extensive areas of cell-rich connective tissue stroma containing bands of osteoid matrix and anastomosing immature bone trabeculae (long arrows) (original magnification 40×). (b) Scattered clusters of multinucleated giant cells (short arrows) (original magnification 40×). (c) Plump osteoblasts can be seen surrounding the interconnecting immature bony trabeculae. Cellular osteoid (star) and some myxomatous areas (polygon) can also be detected (H&E stain, magnification 100×).\n\nFollowing treatment, the patient was administered 50 mg diclofenac potassium tablets through oral administration, three times daily for one week. The patient responded well to both the surgery and post-operative treatment. The healing was uneventful and no worrying findings were observed through the 6-week follow up.\n\n\nDiscussion\n\nCentral giant cell granuloma (CGCG) is a benign intra-bony lesion that was first introduced to the medical literature by Jaffé in 19538,9. It is a pathology that more commonly affects young females. Some authors like Chuong et al.10 and Ficarra et al.11 tend to classify it into aggressive and non-aggressive subcategories, due to varying clinical courses of reported cases12. Similarly, it can represent varying radiological pictures ranging from undulated unilocular to multilocular radiolucencies13.\n\nJOF is an aggressive, non-odontogenic, benign tumor of the bone14. It is an entity that affects individuals younger than 15 years of age with no sex predilection15. Radiographically, it is shown as a well-circumscribed unilocular or multilocular radiolucency that may exhibit ground-glass or even more discrete opacifications7,16.\n\nThe term “hybrid lesions” was first introduced by Waldron and El-Mofty during their work on cases of ameloblastoma17. Later, it was generally accepted that “hybrid” lesions represent those lesions that comprise significant portions of histopathological features of multiple, completely different entities18–20. We believe that this reported case represents a hybrid lesion. Previous cases of hybrid CGCGs associated with different fibro-osseous entities have been reported in the literature. Until now, to our knowledge ten cases of hybrid lesions have been documented. Out of them, five cases were of CGCG associated with ossifying fibroma, three cases of CGCG with fibrous dysplasia and one case was of CGCG with cemento-osseous dysplasia21. To our knowledge, excluding this case, only one analogous case has been reported22.\n\nAs regards to the histopathological picture of the two lesions, In some instances, CGCG can show confined peripheral areas of reactive bone and osteoid formation in its highly vascular and cellular connective tissue stroma, as indicated by De Corso et al.9,23 Moreover, in juvenile ossifying fibroma, multinucleated giant cells and areas of hemorrhage can be found scattered in its sometimes hypercellular fibrous matrix7, characteristics that can undoubtedly confuse investigators upon making a diagnosis and accordingly, choosing a treatment plan. An additional fact that may further complicate case management is the presence of hybrid lesions.\n\nThe association of giant cells with fibro-osseous lesions could be explained on the basis of different postulations made by many authors. In the case report of Shetty et al.24, the authors state it may just be a coincidental finding or correlated by representing a “reactive process rather than a feature of a separate lesion”. While Farzaneh et al.25 and Penfold et al.26 assume that “osteoblasts may activate osteoclast-type giant cells through paracrine mechanisms”27. Accordingly, a debate has risen between two opinions, whether JOF had developed in a more central position first and then initiated an adjacent giant cell reaction or whether two lesions simply originated independently.\n\nConsidering this specified case, we were startled by the prominent divergence in the course of the lesion; i.e., its response to the selected treatment option. It raised multiple questions and so a decision was made to put the intrigue at ease by embarking on research to find answers. It was firstly suspected that there was a concealment of some histopathological features in the previous biopsy that could direct us to accurate deductions on such a lesion. This instinct was fortunately confirmed. By returning to the histopathological slides of the previous incisional biopsy, very minimal marginal areas of the lesion did indeed contain areas of fibrous stroma with few immature bony trabeculae, which led to the thought that they could in fact be indicative of juvenile ossifying fibroma, of the trabecular subtype, rather than what was previously thought to have been regions of an inflammatory bodily reaction in the form of ossification. Bearing that deduction in mind, along with the results of the recent biopsy, and correlating this to the clinico-radiographical information obtained, the case was identified as a hybrid lesion of both CGCG and JTOF.\n\nOwing to the small number of reported hybrid lesions, their biologic behavior cannot be predicted and hence their treatment is uncertain. As for the chosen treatment in our case, no other surgical intervention was made in an attempt by the surgeon to be as conservative as possible considering the young age of the patient. The surgeon also refrained to prescribe any pharmacological agents with the exception of analgesics three times daily for one week after surgery. The patient was given the appropriate self-care instructions following surgery in front of her guardian and the latter was asked to report the patient to the operating surgeon for follow up. The healing was uneventful for six weeks, but the patient was unfortunately lost for the planned three, six and twelve-month follow up.\n\nTo conclude, in our search of the literature no cases of similar hybrid lesions were reported, except for one reported by Geetha et al.22 To our knowledge, our case is the second to be documented as a hybrid lesion involving CGCG and JOF. Therefore, we believe that this report would significantly contribute to the understanding and clinical management of such a rare lesion.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\n\nConsent\n\nWritten informed consent for publication of clinical details and clinical images was obtained from the guardian of the patient.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nRegezi JA, Sciubba JJ, Jordan RCK: Oral Pathology. Fourth ed. United States of America: SAUNDERS; 2003.\n\nJaffé HL: GIANT-CELL REPARATIVE GRANULOMA, TRAUMATIC BONE CYST, AND FIBROUS (FIBRO-OSSEOUS) DYSPLASIA OF THE JAWBONES. Paper presented at: Monthly conference of the New York Institute of Clinical Oral Pathology; New York, 1952.\n\nEl-Naggar Ak, Chan JKC, Takata T, et al.: WHO Classification of Head And Neck Tumours. 4th ed. Lyon: International Agency for Research on Cancer; 2017. Reference Source\n\nEversole LR: Clinical outline of oral pathology: diagnosis and treatment. China, 2011. Reference Source\n\nMartinez A: Benign tumors / tumor-like conditions: juvenile ossifying fibroma. In. PathologyOutlines.com. 2018.\n\nAndersen LF, Philipsen Ole, Hans P: Oral giant cell granulomas. A clinical and histological study of 129 new cases. Acta Pathol Microbiol Scand A. 1973; 81A(5): 606–16. PubMed Abstract \| Publisher Full Text\n\nNeville BW, Allen CM, Damm DDC, et al.: Oral and Maxillofacial Pathology. Fourth Edition ed. Canada: ELSEVIER; 2016. Reference Source\n\nEhtisham M, Kaur P, Nissar S, et al.: Central Giant Cell Granuloma: Uncommon Yet Important. Journal of Dental and Medical Sciences. 2016; 15(5): 74–80. Publisher Full Text\n\nGulia M, Sood N: Giant Cell Reparative Granuloma Presenting as Cheek Swelling: Case Series of Four Patients. Otolaryngology. 2016; 6(5): 262. Publisher Full Text\n\nChuong R, Kaban LB, Kozakewich H, et al.: Central giant cell lesions of the jaws: a clinicopathologic study. J Oral Maxillofac Surg. 1986; 44(9): 708–713. PubMed Abstract \| Publisher Full Text\n\nFicarra G, Kaban LB, Hansen L: Central giant cell lesions of the mandible and maxilla: A clinicopathologic and cytometric study. Oral Surg Oral Med Oral Pathol. 1987; 64(1): 44–49. PubMed Abstract \| Publisher Full Text\n\nJadu FM, Pharoah MJ, Lee L, et al.: Central giant cell granuloma of the mandibular condyle: a case report and review of the literature. Dentomaxillofac Radiol. 2011; 40(1): 60–64. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWeidner N, Cote RJ, Suster S: Oral Cavity and Jaws. DeFrancesco K, ed. Modern Surgical Pathology. 2nd ed. 2009.\n\nLohe VK, Degwekar SS, Bhowate RR, et al.: Rapidly maturing juvenile ossifying fibroma: a case report. Dentomaxillofac Radiol. 2011; 40(3): 195–198. PubMed Abstract \| Publisher Full Text\n\nWilliams HK, Mangham C, Speight PM: Juvenile ossifying fibroma. An analysis of eight cases and a comparison with other fibro-osseous lesions. J Oral Pathol Med. 2000; 29(1): 13–8. PubMed Abstract \| Publisher Full Text\n\nKamble KA, Guddad SS, Guddad SS, et al.: Central giant cell granuloma: A case report with review of literature. Journal of Indian Academy of Oral Medicine Radiology. 2016; 28(1): 98–101. Publisher Full Text\n\nWaldron CA, el-Mofty SK: A histopathologic study of 116 ameloblastomas with special reference to the desmoplastic variant. Oral Surg Oral Med Oral Pathol. 1987; 63(4): 441–451. PubMed Abstract \| Publisher Full Text\n\nSeifert G, Donath K: Hybrid Tumours of Salivary Glands. Definition and Classification of Five Rare Cases. Eur J Cancer B Oral Oncol. 1996; 32B(4): 251–259. PubMed Abstract \| Publisher Full Text\n\nIde F, Horie N, Shimoyama T, et al.: So-called Hybrid Odontogenic Tumors: Do they really exist? Oral Pathology and Medicine. 2001; 6(1): 13–21. Publisher Full Text\n\nChoudhari SK, Gadbail AR: Hybrid odontogenic tumors: a controversy. In Pathol Oncol Res. 2015; 21(2): 501–2. PubMed Abstract \| Publisher Full Text\n\nJawanda MK, Narula R, Shankari M, et al.: Hybrid lesions comprising central giant cell granuloma and fibrous dysplasia: A diagnostic challenge for pathologist. J Oral Maxillofac Pathol. 2015; 19(3): 408. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGeetha NT, Pattathan RK, Shivakumar HR, et al.: Fibro-osseous lesions vs. central giant cell granuloma: A hybrid lesion. Ann Maxillofac Surg. 2011; 1(1): 70–3. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDe Corso E, Politi M, Marchese MR, et al.: Advanced giant cell reparative granuloma of the mandible: radiological features and surgical treatment. Acta Otorhinolaryngol Ital. 2006; 26(3): 168–172. PubMed Abstract \| Free Full Text\n\nShetty K, Giannini P, Leigh J: A hybrid giant cell granuloma and fibro-osseous lesion of the mandible. Oral Oncology. 2004; 40(6–7): 81–84. Publisher Full Text\n\nFarzaneh AH, Pardis PM: Central giant cell granuloma and fibrous dysplasia occurring in the same jaw. Med Oral Patol Oral Cir Bucal. 2005; 10 Suppl 2: E130–E132. PubMed Abstract\n\nPenfold CN, McCullagh P, Eveson JW, et al.: Giant cell lesions complicating fibro-osseous conditions of the jaws. Int J Oral Maxillofac Surg. 1993; 22(3): 158–162. PubMed Abstract \| Publisher Full Text\n\nCrusoé-Rebello I, Torres M, Burgos V, et al.: Hybrid lesion: central giant cell granuloma and benign fibro-osseous lesion. Dentomaxillofac Radiol. 2009; 38(6): 421–425. PubMed Abstract \| Publisher Full Text" }	[ { "id": "51869", "date": "15 Aug 2019", "name": "Maedeh Ghorbanpour", "expertise": [ "Reviewer Expertise molecular pathology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn keywords: Fibrous dysplasia should be deleted.\n\nPlease check the keywords to be according to MeSH system.\n\nIn introduction: Last sentence of second paragraph about JTOF? In most references psammomatoid type of the JOF is more common in craniofacial skeleton than trabecular type\n\nFigure 1: It would be better to add panoramic view or CBCT of the lesion.\n\nFigure 6C: It would be better to use higher magnifications for showing osteoblasts and cellular osteoid.\n\nThe first sentence of the “Discussion” needs appropriate citation. The original citation needs to be applied to address Jaffé, 1953 original article.\n\nIs this case the first hybrid case of CGCG and JTOF? If not, please provide a literature review of the same cases in a table.\n\nIt is recommended to summarize previously reported cases of hybrid CGCG in a separate table.\n\nThe manuscript needs major revision.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Partly\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Partly\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [] }, { "id": "52552", "date": "18 Sep 2019", "name": "Suwarna B. Dangore-Khasbage", "expertise": [ "Reviewer Expertise Oral Medicine and Oral Radiology" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAuthors have described an uncommon case of Introduction is relevant, adequate and properly written. a case report is described in proper sequence with investigation and diagnosis with justification. Enough details are provided about the case.\n\nSufficient discussion is included depicting the importance of the findings and their relevance to future understanding of disease processes The overall case report is interesting and useful to readers.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [] }, { "id": "54099", "date": "07 Oct 2019", "name": "Rania H Younis", "expertise": [ "Reviewer Expertise Oral and maxillofacial pathology and tumor immunology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\n(JTOF) The authors present a hybrid case report of CGCG, with juvenile trabecular ossifying fibroma.\nThe authors should mention the soft ware used to do the surgical planning (Figure 1), was it a CBCT?\nIt is recommended that the authors provide a high power representative of the heavily cellular stroma and pseudoaneurysmal spaces, characteristic of the Juvenile trabecular ossifying fibroma (JTOF).\nAdd reference to support the triamcinolone acetonide injection.\nIn the discussion the authors need to reflect more on the differential diagnoses, that the bone deposition may be part of a reparative granuloma histologically, yet the rapid clinical expansion speaks against a reparative process.\nThat JTOF can be associated with MNGCs, yet the incisional biopsy had enough evidence of a CGCG component to support a hybrid lesion.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Partly\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-1218
https://f1000research.com/articles/8-583/v1	29 Apr 19	{ "type": "Research Article", "title": "Turning the tables: A university league-table based on quality not quantity", "authors": [ "Adrian G. Barnett", "David Moher", "David Moher" ], "abstract": "Background: Universities closely watch international league tables because these tables influence governments, donors and students. Achieving a high ranking in a table, or an annual rise in ranking, allows universities to promote their achievements using an externally validated measure. However, league tables predominantly reward measures of research output, such as publications and citations, and may therefore be promoting poor research practices by encouraging the “publish or perish” mentality. Methods: We examined whether a league table could be created based on good research practice. We rewarded researchers who cited a reporting guideline, which help researchers report their research completely, accurately and transparently, and were created to reduce the waste of poorly described research. We used the EQUATOR guidelines, which means our tables are mostly relevant to health and medical research. Results: Our cross-sectional tables for the years 2016 and 2017 included 14,408 papers with 47,876 author affiliations. We ranked universities and included a bootstrap measure of uncertainty. We clustered universities in five similar groups in an effort to avoid over-interpreting small differences in ranks. Conclusions: We believe there is merit in considering more socially responsible criteria for ranking universities, and this could encourage better research practice internationally if such tables become as valued as the current quantity-focused tables.", "keywords": [ "meta-research", "research quality", "research reporting", "league tables" ], "content": "Introduction\n\nLeague tables are used by universities to advertise their value, recruit staff and students, and attract funding, particularly philanthropic funding. There are many international league tables including the Times Higher Education World University Rankings, QS World University Ranking and CWTS Leiden Ranking. There are also national league tables, such as the Complete University guide in the UK, but in this study we only consider international league tables. We also focus on research, and so we do not consider league tables or criteria that focus on teaching or service. Many universities have dedicated web pages that promote their league table rankings with news stories and graphics1–3. League tables create opportunities for universities to write positive stories based on either: i) their ranking, or ii) a large rise in their ranking as the tables are updated annually. Rankings can also be stratified by country, scientific field, or the league table’s criteria (e.g. teaching or research), offering multiple opportunities for positive stories. The league tables are made by groups that are independent of universities, and therefore give an external marker of quality.\n\nExample quotes from university web pages concerning their position in league tables are below and these demonstrate some of the ways universities use league tables for self-promotion.\n\n“The University’s outstanding performance in the Leiden Ranking sent a strong signal to potential partners and collaborators that top-quality, highly cited research was produced across all disciplines.” http://tinyurl.com/y94tomgr\n\n“Deakin has climbed 62 places to enter the world’s top 300 universities, according to the latest prestigious QS World University Rankings [...] The latest ranking places Deakin in the top 1.1 per cent of universities in the world.” https://tinyurl.com/y9xzmtpk\n\n“The University of Toronto is among the best universities in the world for graduate employability, a new independent study says.” https://tinyurl.com/ydxju5xu\n\n“These results demonstrate that the University of Toronto is a consistent producer of impactful, world-class research across a broad range of disciplines” https://tinyurl.com/yd3uz83m\n\nThe quotes were selected to illustrate how universities value league tables. They were found by selective searching and are not a representative sample.\n\nUniversity managers often want to maintain a high ranking or increase their ranking in international league tables, and may implement top-down policies that encourage their staff to work in ways that will achieve this. A review of the impact of university league tables in the UK found that they, “appear to be having a significant influence on institutions’ actions and decision-making”4. These changes to research practices may have societal costs. For example, encouraging researchers to focus on quantity so that rankings based on publications numbers increase, may lead researchers to cut corners in order to increase their output at the expense of quality5.\n\nLeague tables could potentially be used to promote positive changes in research culture if they included criteria of good research practice, which might then encourage university managers to widely promote good practice.\n\nThe International Ranking Expert Group (IREG) audit university league tables and aim to strengthen public awareness and understanding of university rankings. A recent inventory by IREG found 17 international league tables6, although two are based solely on web traffic and one concerns environmental sustainability. Of the remaining 14 tables, 12 use publication numbers, and 12 use citations.\n\nAlthough papers and citations are commonly used, every league table uses their own method to count them. Variations include:\n\nOnly papers or citations from selected “high quality” journals\n\nOnly relatively highly cited papers\n\nOnly papers cited by industry\n\nCitation numbers divided by the number of papers\n\nPaper numbers divided by the number of staff\n\nThe differences between league tables could reflect genuine differences of opinion in the best way to use the data. It could also be somewhat due to a desire by league tables to differentiate themselves and so produce novel results. It could also be because papers and citations are imperfect proxies of quality, and so there are multiple opinions on how best to refine them.\n\nA seminal paper on institutional ranking (including hospitals and schools) in 1996 by Goldstein and Spiegelhalter stated that responsible rankings, “may provide relevant information to universities, students, funders and governments”7. However, they also cautioned about the need to consider data quality, uncertainty in the rankings, gaming by institutions, and unwarranted conclusions based on small changes in ranks. A report on the use of public league tables recommended that every table should have an appropriate and prominent “health warning” about their limitations8.\n\nThe criteria used by university league tables have been criticised for lacking construct validity9 and for experiencing implausibly large changes from year to year10, some of which were due to calculation errors and methodological changes11.\n\nA report on the use of citation statistics warned that “citation data provide only a limited and incomplete view of research quality”12. An analysis of misprints in citations suggested that most researchers simply copy citations without reading the actual paper13, which undermines their face validity as a ranking criteria. Citations and paper numbers can be gamed14,15, and gaming by researchers can greatly alter a university’s ranking11. Concerns about the misuse of simplistic metrics in research led to the Leiden Manifesto in 2015, which set out ten principles for the proper use of metrics for evaluating researchers and institutions16. In 2017 the Leiden group created ten more principles for responsibly ranking universities17, which included transparency and acknowledging the uncertainty in rankings.\n\nTo our knowledge, no current international league table includes a measure of best publication practices, by which we mean established methods that increase the robustness, transparency and reproducibility of research. There is an international league table of potentially questionable research practice, which is the Retraction Watch table of individual researchers ranked by their number of retracted papers18.\n\nExamples of good research practice are:\n\nIncluding key stakeholders in forming research questions19\n\nPublishing a protocol and ensuring that the results presented match those planned in the protocol20\n\nPublishing results even when they are statistically negative or potentially commercially damaging21\n\nUsing reporting guidelines to write-up the results22\n\nSharing data and code where available21\n\nUnlike the traditional metrics, such as the number of publications, used by current league tables, these metrics are prerequisites to solving recognised problems in science. Recent evidence points to a growing reproducibility crisis in many fields of research, which is only possible to examine when sharing of data, code, materials and methods takes place.\n\nGood research practices help reduce research waste, which can occur when researchers cut corners in order to progress in the “publish or perish” game. Avoidable research waste is an enormous problem and an estimated 85% of the current investment in health and medical research is wasted due to poor research practice, which is billions of dollars per year23.\n\nIn this paper we examine one of these good research practices by examining when authors cited an EQUATOR reporting guideline24. EQUATOR stands for: Enhancing the QUAlity and Transparency Of health Research, and they are a wide-ranging suite of more than 400 guidelines that cover every common research study design. There is evidence that using a reporting guideline improves the quality of the published paper25,26. Our key assumption is that citing the guideline is an indicator of good research practice. An important difference from our approach compared with previous league tables, is that we reward the universities whose researchers give the citation, not the universities of researchers who receive the citation.\n\nThere are four EQUATOR centres around the world (UK, France, Canada and Australasia) with the aim of promoting the use of the guidelines worldwide. Many of the most commonly used EQUATOR guidelines have been translated into multiple languages.\n\nThere is a wide literature on rankings and university league tables including discussions of policy27, design28 and statistical critiques7, as well as systematic reviews29 and books30. We do not review this literature in detail, as our primary aim was to identify whether a league table could be constructed based on good research practice.\n\n\nMethods\n\nWe use the phrase “university rankings” to be consistent with the existing league tables. However, “institutional rankings” would be more accurate because we include research institutes that may be affiliated with universities but do not graduate students, such as the “Baker Heart and Diabetes Institute”.\n\nWe counted papers that cited one of the EQUATOR guidelines for clinical trials (CONSORT)31, systematic reviews (PRISMA)32, and observational studies (STROBE)33. We chose these three guidelines because they cover three commonly used study designs. Each guideline was published simultaneously across multiple journals, which was done to increase their reach into multiple fields. We therefore counted citations to any of the original papers or updates to the guidelines (see Supplementary List 1)34. If a paper cited multiple EQUATOR papers, then only one was counted.\n\nTo include only papers that adhered to the first item on the CONSORT and PRISMA guideline check-lists, which is to include the study design in the title, we only included papers that included the following in their title:\n\nFor CONSORT papers: “randomised trial” OR “randomized trial” OR “RCT”\n\nFor PRISMA papers: “systematic search” OR “systematic review” OR “systematic literature review” OR “scoping review” OR “meta-analyses” OR “meta-analysis” (including versions without hyphens)\n\nWe did not include a restriction for STROBE papers because there are many observational study designs and any list we created might exclude valid papers.\n\nTo focus on original research, we included publication types of Articles or Reviews, and excluded Editorials, Commentaries and Corrections.\n\nWe aimed to sum citations per year and we examined the two most recent complete years of data by using papers published in 2016 or 2017.\n\nWe used Scopus to identify citations because it is a recognised database for citations that is used by three international league tables, and because of the ease of extracting the data using the rscopus package in R35 (Version 0.6.3). We used the rentrez package in R (version 1.2.1) to extract meta-data on the papers from Pubmed36. Papers were excluded if they did not have a digital object identifier (DOI), because this was the key linking variable for extracting the affiliation data. The data extraction from Scopus was performed on 19 December 2018.\n\nWe extracted all authors’ countries and affiliations. The affiliation data is free text and required extensive cleaning to extract a standardised set of universities. Affiliations were changed to:\n\nRemove departments, for example, “Mansoura University, Urology and Nephrology Center” to “Mansoura University”\n\nInclude non-Roman letters, for example, “Universite de Montreal” to “Université de Montréal”.\n\nRemove locations, for example: “Massey University, Auckland” to “Massey University”. The exception was where the location was needed to differentiate the university, for example the University of Newcastle in the UK and Australia.\n\nRemove unnecessary prefixes, for example: “The University of Sydney” to “University of Sydney”\n\nSpell-out acronyms, for example: “UCL” to “University College London”\n\nConsolidate dual names, for example: “University of Reykjavik” to “Reykjavik University”\n\nConsolidate institutes associated with a university, for example: “The Ottawa Hospital” is associated with the “University of Ottawa”. We used the list of 1,802 affiliated institutions provided by the 2018 Leiden ranking17.\n\nWe changed vague affiliations to missing, for example “Faculty of Health”.\n\nWe standardised affiliations to ensure that citations were consolidated into a single university rather than being split over two or more universities and hence creating a falsely low position in our league table.\n\nA flow chart of the data collection and management is in Supplementary Figure 134.\n\nTo create a score per university, we summed the total number of citing papers per university per year. To better divide the credit from a citation, we used an organisational-level fractional count of author affiliations per paper37. So, for example, if a paper had two affiliations in the address list, one from Queensland University of Technology and one from Ottawa Hospital Research Institute, then each university would gain 0.5. A fractional count avoids the situation where universities gain a full point even when their staff member was only one of multiple authors.\n\nWe examined the amount of missing affiliation data by country to look for biases in the affiliation data that may disadvantage particular universities or geographic regions in our league table. We also included “Missing” as a separate university, in order to show the relative importance of missing data.\n\nWe accounted for uncertainty in our league table using a bootstrap procedure38. We randomly resampled with replacement from all the citing papers and recalculated each university’s score and rank. We repeated this resampling 1,000 times. To summarise this uncertainty we created a bootstrap 95% confidence interval for the rank.\n\nWe examined changes over time by comparing the ranks of universities in the top 200 in 2016 and 2017. We used a Bland–Altman plot to examine how ranks changed between these two years39. For comparison, we also used a Bland–Altman plot of the THE World University Rankings using their research criterion, which combines a reputation survey, data on research income and paper numbers40.\n\nWe qualitatively self-assessed our league table against the ten principles for responsible ranking from the Leiden group41.\n\nAs a comparison to our good research practice table, we created a standard league table based on counting each university’s papers for the years 2016 and 2017. We counted articles only, not books, editorials or letters. To match our good practice table which is focused on health and medical research, we only included papers in the three subject areas of Dentistry, Health Professions and Nursing. These data were from Scopus.\n\nWe present our results as a table using the total score per university per year and give an integer rank to universities in each year. This implies a monotonic order, where each university performed better than the university below it. This is unlikely to be true, and to give a better impression of performance we used clustering to group universities into five clusters. We chose five as an a priori opinion of the number of meaningful clusters. We used a Bayesian clustering model defined as:\n\n\n\nwhere S(i, t) is our score for university i in year t. The five cluster means (x¯) are ordered from low to high. For each university we estimate their cluster, c(i, t) ∈ c(1, 2, 3, 4, 5), which comes from a categorical distribution with five probabilities π(1), . . . , π(5). These probabilities came from the sum of five uniform prior distributions which were formulated so that the minimum probability for each cluster was 1% (π ≥ 0.01). This was an attempt to avoid small clusters of just a few universities. We only applied the clustering algorithm to universities with a score of 2 or above, which removed the large number of universities with small samples sizes and low scores. We cross-tabulated the median clusters by year to show how many universities changed between 2016 and 2017.\n\nThe data extraction and analyses were made using R version 3.5.242. The clustering model was fitted in WinBUGS (version 1.4.3)43 and we visually checked the mixing of the Markov chain Monte Carlo estimates. The data and code that created the tables is available here: https://github.com/agbarnett/league.tables.\n\nIn summary, the aim of our table was to score universities using the EQUATOR guidelines, with higher scores indicative of better research practice. We also included measures of uncertainty via the bootstrap and attempted to cluster similar universities. We report our results using the STROBE guidelines33.\n\n\nResults\n\nOur tables included 14,408 papers giving a total of 47,876 author affiliations that could be counted. The average number of affiliations per paper was 3.3.\n\nThe number and percent of missing affiliation data are shown by country in Table 1. If the country was missing then the affiliation was also likely to be missing. The most amount of missing data was in the USA. Overall the percent of missing affiliation data was small, at just 0.5% of all affiliations.\n\n“Missing” is included as a nominal country, that is the affiliation and country data were both missing. Countries ordered by number missing.\n\nBefore examining institutions, we first examine the scores by regions and countries, and the top ten regions and countries are shown in Table 2. The rank order of the top ten was the same for the regions and countries, except for the tenth ranked country, which was Denmark in 2016 and Spain in 2017. Every region and country in the top ten had a higher total score in 2017 than 2016, reflecting an increased use of the EQUATOR guidelines. The highest ranking regions and countries in the table are familiar producers of research.\n\nThese results exclude “Missing” as a nominal country or region.\n\nThe top ten universities in each year are in Table 3. We have presented the scores in this paper to one decimal place, but would use rounded integers in public tables to discourage readers over-interpreting small differences. The University of Toronto had the highest score for papers citing the EQUATOR guidelines in both years. Although the proportion of missing affiliation data in the entire data set is small (just 0.5%), “Missing” was in the top ten in both years.\n\nUniversities are ordered by their score in each year. The cluster column is the median cluster from the Bayesian model, with ‘5’ the highest cluster. The rank is the median rank and 95% bootstrap confidence interval in brackets. The standard rank is based on counting each university’s annual papers.\n\na There was no standard rank for missing affiliations. b Tied.\n\nThe University of Toronto was ranked highest for good research practice in both years, and there was little uncertainty in this top ranking as the bootstrap confidence intervals were rank 1 to 2 in 2016 and rank 1 to 1 in 2017. The University of Sydney was ranked second in both years.\n\nThe clustering model selected only a small number of universities to be in the highest category of ‘5’, despite our attempt to avoid small clusters by formulating a minimum prior probability of 1%. Summary statistics for the five clusters are in Supplementary Table 134.\n\nThere was relatively little movement in clusters between years for the best clusters of ‘3’ to ‘5’ (Table 4). There was more movement over time between the lowest two clusters of ‘1’ and ‘2’. Only two universities moved by two or more clusters, which was from ‘1’ to ‘3’.\n\nThe diagonal numbers in bold correspond to no change from 2016 to 2017. ‘5’ is the highest cluster with the best score.\n\nThe 95% bootstrap intervals were wider for universities outside the top ten. For example, for the university ranked 100 in 2017, the 95% interval was from rank 63 to 176. The width of the interval increased by an average of 13.6 for every 10 increase in rank (95% CI 13.0 to 14.1 using linear regression; see Supplementary Figure 234). This increase was due to the reduced sample size (number of papers) for lower ranked universities.\n\nThe universities in our top 10 had varied results using a standard ranking, with some being in the top 10 and others outside the top 100. Two Chinese universities ranked in the top ten in our good research practice ranking, but were outside the top 100 using the standard table. Erasmus University and The University of Ottawa also did much better on the good research practice ranking that the standard ranking. The Spearman’s rank correlation between the standard ranking and our good practice ranking was 0.59.\n\nComplete tables for all universities with a score of two or above are available online: https://aushsi.shinyapps.io/equator (available until 2020). These interactive tables allow examination of the results by year, geographical region and selected countries. The top 50 universities per year are shown in Supplementary Tables 2 and 334.\n\nWe show the agreement in university ranks between years using Bland–Altman plots in Figure 1. For both our league table and the THE table, there was less change in the highest ranking universities, and more movement between years at lower ranks. The Bland–Altman limits of agreement were –60 to 60 in our table and –46 to 43 for the THE table.\n\nWe only examine universities in the top 200 in both years, which is 161 in our table and 184 in the THE table. The dashed horizontal lines are the Bland–Altman limits of agreement.\n\nWe assessed our Good Research Practice league table against the ten Leiden principles in Table 5.\n\n\nDiscussion\n\nCurrent league tables place a high value on the quantity of research outputs and citations. The irony is that the biomedical literature is littered with publications that cannot be reproduced, have substantive reporting biases and mistakes in study design, making much of such output unusable20. It is hard to imagine why most universities continue to support the current ranking schemes given that they may be reducing the positive value universities have on society. We believe there is merit in considering alternative more socially responsible criteria for ranking universities.\n\nWe have created a league table based on a good research practice criterion that shows which universities are performing well and which could improve. We aimed to include all eligible universities, and so our results should be inclusive and generalisable.\n\nLindner et al recently examined whether metrics and incentives could be developed to encourage scientists to use high-quality methods and publish “negative” studies44. They concluded that, “If rigorous, innovative studies of significant issues and publication of valid, reproducible results are desired, the best way to achieve those objectives is to explicitly evaluate and reward scientists based on those criteria.”\n\nLane suggested that new metrics should capture “the essence of what it means to be a good scientist”45 and future league tables could include:\n\nthe percent of papers that are open access (as suggested by Nichols and Twidale46),\n\npapers where the data and/or code have been openly shared,\n\nstudies that were pre-registered and published in a timely manner,\n\npapers with a published protocol.\n\nHowever, league tables generally rely on large volumes of data to create scores, meaning these criteria would need to be automated. At present we could only likely automate whether matching data or protocol paper existed, and not whether the data was complete or whether the authors followed the protocol. Detailed data that cannot be automated can be collated on a smaller scale using audits47,48.\n\nWe could expand our criteria to include more of the EQUATOR guidelines, such as the STARD guidelines for diagnostic accuracy studies49. Including more EQUATOR guidelines would increase the sample size per university and so would likely reduce some of the variation between years shown in Figure 1.\n\nWe did not adjust for the size of the university to produce a relative measure of performance. Hence our table is biased towards larger universities that have more staff, an issue recognised by the Leiden manifesto on metrics17. An ideal standardisation would be to adjust for the number of papers that failed to cite an EQUATOR guideline when appropriate. This could be used to give an indication of performance regardless of size, and would also show the potential improvement for each university.\n\nOne surprising result from our tables was the high rank of “Missing”. This shows the importance of correctly completing affiliations, and universities could increase their rankings (in our table and others) by promoting a clear and consistent affiliation to their staff. We recommend that all league tables report the amount of missing data and show its ranking in their tables. We also recommend, as have others7,17, that all league tables include a measure of ranking uncertainty.\n\nThere are many limitations to constructing a university league table, and our tables should be treated as suggestive rather than definitive7.\n\nIt is impossible to numerically validate our table because there is no gold standard ranking against which we can compare our results. We qualitatively assessed our own performance against the ten Leiden principles, but others may be more critical.\n\nA valid concern with our table is that it would be gamed, with researchers simply citing an EQUATOR guideline without engaging with it. This is very likely to happen, but we cannot estimate the scale of this problem. This is less likely in journals that appropriately implement reporting guidelines because there is an internal check. The harms from such gaming could be outweighed by the number of researchers and universities that genuinely engage with the EQUATOR guidelines. Benefits would likely include greater awareness of the guidelines, and prompting researchers who were already aware of them to use them more rigorously. Complete and transparent reporting has been indicated as an essential prerequisite in dealing with the reproducibility crisis50. Some token engagement with a guideline could be spotted by the paper’s peer reviewers, although peer reviewers often have limited time and have an imperfect record of spotting mistakes in papers51. It may be possible to automate how the paper has adhered to the guidelines and produce a report that is shared with the authors, reviewers and editor(s), and there is an ongoing trial at the journal BMJ Open of such a tool52.\n\nThe free text affiliation data from Scopus were challenging to process as they were often incomplete and inconsistent. Some universities have multiple versions of their name, including acronyms and English-language versions. We made extensive searches and asked international colleagues to check where consolidations could be made. However, we are very likely to have missed some consolidations, and hence some universities may be too low in our tables because their data has been spread across multiple names.\n\nWe tried to examine a correlation in ranks between our tables and those of the Times Higher Education World University Rankings and CWTS Leiden Ranking. However, it was very difficult to correctly merge the data because of the large variation in affiliation names. Just one of many examples is we use “Mayo Clinic”, whereas the Times Higher Education uses “Mayo Medical School”, and this institute is not included in the CWTS Leiden Ranking.\n\nWe could only find one previous related study, which was an international ranking that aimed to measure research quality by using membership on academic editorial boards of professional journals53. They extracted researchers’ names from the websites of 115 economics journals creating a sample of over 3,700 researchers, and created league tables of researchers and universities. Their conclusion was that their table could be used to find experts to evaluate research quality.\n\n\nConclusions\n\nInternational league tables are fuelling a hyper-competitive research world that values quantity over quality. We attempted to create the first international league table that focused on good research practice. This is part of a long recognised need to focus on quality over quantity, which was raised by Doug Altman in 1994 when he said, “We need less research, better research, and research done for the right reasons”54. Our table is not a perfect measure of research quality, but we hope that such tables will become valued by right-thinking universities whose goal should be to produce robust research rather than simply the most amount of research.\n\n\nData availability\n\nA random selection of 500 rows of the data has been made available (see below). The public sharing of data for the purpose of reproducibility with a specific party is permissible upon written request and explicit written approval and the dataset remains with the customer/research. Requests can be made to: integrationsupport@elsevier.com. Zenodo: agbarnett/league.tables: Ready for journal submission. https://doi.org/10.5281/zenodo.259401634.\n\nZenodo: agbarnett/league.tables: Ready for journal submission. https://doi.org/10.5281/zenodo.259401634.\n\nSupplementary List 1. List of papers for which citations were counted.\n\nSupplementary Figure 1. Flow chart of the data collection and management steps.\n\nSupplementary Table 1. Summary statistics for the five clusters from the Bayesian model. Estimated probability for each cluster (π), mean scores x¯ , and 95% credible intervals for means.\n\nSupplementary Table 2. Top 50 ranked universities in 2016.\n\nSupplementary Table 3. Top 50 ranked universities in 2017.\n\nSupplementary Figure 2. Scatter plot of the width of the 95% bootstrap interval against rank using the top 200 universities in both years.\n\nWhere appropriate, extended data are held under the MIT License.\n\n\nSoftware availability\n\nSource code used for analysis available from: https://github.com/agbarnett/league.tables.\n\nArchived data and code at time of publication: https://doi.org/10.5281/zenodo.259401634.\n\nLicence: MIT License", "appendix": "Author contributions\n\n\n\nAB and DM conceived the original idea and designed the league table. AB did the data extraction and statistical analyses. AB and DM interpreted the results. AB wrote the first draft of the paper with input from DM.\n\n\nGrant information\n\nA.B. receives funding from the Australian National Health and Medical Research Council (APP1117784).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nQueensland University of Technology: Rankings. 2018. Reference Source\n\nThe University of Queensland: Rankings. 2018. Reference Source\n\nUniversity College London: About UCL. 2018. Reference Source\n\nLocke W, Verbik L, Richardson JTE, et al.: Counting what is measured or measuring what counts? League tables and their impact on higher education institutions in England. Higher Education Funding Council for England, Bristol, UK. 2008. Reference Source\n\nSmaldino PE, McElreath R: The natural selection of bad science. R Soc Open Sci. 2016; 3(9): 160384. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nInternational Ranking Expert Group: IREG inventory of international university rankings 2014–17. 2018. Reference Source\n\nGoldstein H, Spiegelhalter DJ: League tables and their limitations: statistical issues in comparisons of institutional performance. J R Stat Soc A Stat. 1996; 159(3): 385–443. Publisher Full Text\n\nFoley B, Goldstein H: Measuring success: League tables in the public sector. British Academy. 2013. Reference Source\n\nIoannidis JP, Patsopoulos NA, Kavvoura FK, et al.: International ranking systems for universities and institutions: a critical appraisal. BMC Med. 2007; 5(1): 30. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBookstein FL, Seidler H, Fieder M, et al.: Too much noise in the Times Higher Education rankings. Scientometrics. 2010; 85(1): 295–299. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHolmes R: Searching for the gold standard: The Times Higher Education world university rankings, 2010-2014. Asian Journal of University Education. 2015; 11(2): 1–29. Reference Source\n\nAdler R, Ewing J, Taylor P: Citation statistics. Statistical Science. 2009; 24(1): 1–14. Publisher Full Text\n\nSimkin MV, Roychowdhury VP: Read before you cite. Complex Syst. 2003; 14: 269–274. Reference Source\n\nBiagioli M: Watch out for cheats in citation game. Nature. 2016; 535(7611): 201. PubMed Abstract \| Publisher Full Text\n\nFong EA, Wilhite AW: Authorship and citation manipulation in academic research. PLoS One. 2017; 12(12): e0187394. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHicks D, Wouters P, Waltman L, et al.: Bibliometrics: The Leiden Manifesto for research metrics. Nature. 2015; 520(7548): 429–431. PubMed Abstract \| Publisher Full Text\n\nCWTS Leiden Ranking: Responsible use. 2018. Reference Source\n\nMarcus A, Oransky I: Science publishing: The paper is not sacred. Nature. 2011; 480(7378): 449–450. PubMed Abstract \| Publisher Full Text\n\nChalmers I, Bracken MB, Djulbegovic B, et al.: How to increase value and reduce waste when research priorities are set. Lancet. 2014; 383(9912): 156–165. PubMed Abstract \| Publisher Full Text\n\nIoannidis JP, Greenland S, Hlatky MA, et al.: Increasing value and reducing waste in research design, conduct, and analysis. Lancet. 2014; 383(9912): 166–175. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChan AW, Song F, Vickers A, et al.: Increasing value and reducing waste: addressing inaccessible research. Lancet. 2014; 383(9913): 257–266. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGlasziou P, Altman DG, Bossuyt P, et al.: Reducing waste from incomplete or unusable reports of biomedical research. Lancet. 2014; 383(9913): 267–276. PubMed Abstract \| Publisher Full Text\n\nChalmers I, Glasziou P: Avoidable waste in the production and reporting of research evidence. Lancet. 2009; 374(9683): 86–89. PubMed Abstract \| Publisher Full Text\n\nAltman DG, Simera I: A history of the evolution of guidelines for reporting medical research: the long road to the EQUATOR Network. J R Soc Med. 2016; 109(2): 67–77. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCobo E, Cortés J, Ribera JM, et al.: Effect of using reporting guidelines during peer review on quality of final manuscripts submitted to a biomedical journal: masked randomised trial. BMJ. 2011; 343: d6783. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTurner L, Shamseer L, Altman DG, et al.: Consolidated standards of reporting trials (CONSORT) and the completeness of reporting of randomised controlled trials (RCTs) published in medical journals. Cochrane Database Syst Rev. 2012; 11: MR000030. PubMed Abstract \| Publisher Full Text\n\nDill DD, Soo M: Academic quality, league tables, and public policy: A cross-national analysis of university ranking systems. Higher Education. 2005; 49(4): 495–533. Publisher Full Text\n\nProulx R: Higher education ranking and leagues tables: Lessons learned from benchmarking. Higher Education in Europe. 2007; 32(1): 71–82. Publisher Full Text\n\nVernon MM, Balas EA, Momani S: Are university rankings useful to improve research? A systematic review. PLoS One. 2018; 13(3): e0193762. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHazelkorn E: Rankings and the Reshaping of Higher Education. Palgrave Macmillan UK, 2nd edition. 2015. Publisher Full Text\n\nSchulz KF, Altman DG, Moher D, et al.: CONSORT 2010 statement: updated guidelines for reporting parallel group randomised trials. PLoS Med. 2010; 7(3): e1000251. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiberati A, Altman DG, Tetzlaff J, et al.: The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Med. 2009; 6(7): e1000100. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVandenbroucke JP, von Elm E, Altman DG, et al.: Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): explanation and elaboration. PLoS Med. 2007; 4(10): e297. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBarnett A: agbarnett/league.tables: Ready for journal submission. 2019. http://www.doi.org/10.5281/zenodo.2594016\n\nMuschelli J: rscopus: Scopus Database ’API’ Interface. 2018. Reference Source\n\nWinter DJ: rentrez: an R package for the NCBI eutils API. The R Journal. 2017; 9(2): 520–526. Publisher Full Text\n\nWaltman L, van Eck NJ: Field-normalized citation impact indicators and the choice of an appropriate counting method. J Informetr. 2015; 9(4): 872–894. Publisher Full Text\n\nDavison AC, Hinkley DV: Bootstrap Methods and Their Application. Cambridge Series in Statistical and Probabilistic Mathematics. Cambridge University Press. 1997. Publisher Full Text\n\nBland JM, Altman DG: Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986; 1(8476): 307–310. PubMed Abstract \| Publisher Full Text\n\nTimes Higher Education: World university rankings 2015–2016 methodology. 2018. Reference Source\n\nWaltman L, Wouters P, van Eck NJ: Ten principles for the responsible use of university rankings. 2017. Reference Source\n\nR Core Team: R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria, 2018. Reference Source\n\nLunn DJ, Thomas A, Best N, et al.: WinBUGS – a Bayesian modelling framework: concepts, structure, and extensibility. Stat Comput. 2000; 10(4): 325–337. Publisher Full Text\n\nLindner MD, Torralba KD, Khan NA: Scientific productivity: An exploratory study of metrics and incentives. PLoS One. 2018; 13(4): e0195321. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLane J: Let's make science metrics more scientific. Nature. 2010; 464(7288): 488–489. PubMed Abstract \| Publisher Full Text\n\nNichols DM, Twidale MB: Metrics for openness. J Assoc Inf Sci Technol. 2016; 68(4): 1048–1060. Publisher Full Text\n\nGoldacre B: How to get all trials reported: audit, better data, and individual accountability. PLoS Med. 2015; 12(4): e1001821. PubMed Abstract \| Publisher Full Text\n\nBarnett AG, Zardo P, Graves N: Randomly auditing research labs could be an affordable way to improve research quality: A simulation study. PLoS One. 2018; 13(4): e0195613. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBossuyt PM, Reitsma JB, Bruns DE, et al.: STARD 2015: an updated list of essential items for reporting diagnostic accuracy studies. BMJ. 2015; 351: h5527. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGoodman SN, Fanelli D, Ioannidis JP: What does research reproducibility mean? Sci Transl Med. 2016; 8(341): 341ps12. PubMed Abstract \| Publisher Full Text\n\nSchroter S, Black N, Evans S, et al.: What errors do peer reviewers detect, and does training improve their ability to detect them? J R Soc Med. 2008; 101(10): 507–514. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\ndel Pozo Martin Y: BMJ Open trials Penelope. 2017. Reference Source\n\nFrey BS, Rost K: Do rankings reflect research quality? J Appl Econ. 2010; 13(1): 1–38. Publisher Full Text\n\nAltman DG: The scandal of poor medical research. BMJ. 1994; 308(6924): 283–284. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "47855", "date": "13 May 2019", "name": "Ellen Hazelkorn", "expertise": [ "Reviewer Expertise Higher education policy", "with a particular focus on globalization and university rankings having published and advised on these matters." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper presents an interesting perspective on rankings with particular attention given to quality vs quantity argument. There is an extensive literature on rankings and research; the authors refer to some of the key texts. The issue of citations is arguably more problematic than referenced here; the problem of citations includes self-citation and poor practice as well as whether it is a meaningful measure of quality. However, there is a wider debate around whether citations are a meaningful or appropriate measure of 'impact' given that policy attention is increasingly on impact beyond the academic community.\n\nThe authors use examples of research best practice to ask about the extent to which rankings - which predominantly measure research output - are 'genuinely' concerned about the quality of the research or simply the number of papers or the number of citations, etc. This is an admirable attempt to get behind the rankings.\nAs part of the methodological steps taken, the authors encounter a major default with the research data as evidenced by the way in which authors describe their own institutional/university affiliation. This is a major problem for institutions; on the other hand, it can also be one of the cheapest ways to improve in the rankings simply by cleansing the data as well as ensuring that the institutional data supplied is accurate. There are examples of universities which have changed position, up and down, because of this. Surprising how much institutional data is either inaccurate or indeed 'gamed'. The extent of the cleansing problem revealed is nonetheless staggering.\n\nThe results are particularly interesting, particularly the positioning of China in the country and university rankings. This comes at a time when China is being accused of poor practice, this is a very interesting result. It also reflects the increasing multi polarity of global science. While previous decades saw the EU, Japan and the US dominate, as Leydesdorff, Wagner, & Adams (2013)1 argue, today the number of scientific nations now includes more than 40 nations. This is an interesting finding and may challenge perceptions of the scientific world.\nThe paper gives us food for thought albeit it is unlikely to affect the main rankings - Times Higher Education and QS - or even Shanghai's Academic Ranking of World Universities given the level of complexity. Nonetheless, it asks valid questions about whether what is measured is what we think is measured.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "49068", "date": "04 Jul 2019", "name": "David M. Nichols", "expertise": [ "Reviewer Expertise Open access", "scientometrics" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper explores a new approach for ranking universities through a proxy for good research practice.\n\nA positive aspect of the paper is that rather than simply complaining about the methods of existing ranking systems the authors have attempted to practically provide an alternative. I agree that no other ranking attempts to rank mainly (or entirely) by best practice, and this is a novel contribution of this paper. As such the paper provides a valuable counter-balance to the current dominant ranking methods. Although the paper focuses on the global rankings the same arguments would seem to apply to national ranking systems such as the REF (UK) and the PBRF (New Zealand). It would be good to add a note to this effect.\nOne linkage to prior work which could be added is the discussion around research soundness in Moore et al.1. In discussing “soundness” rather than “excellence” they say: “… our focus should be on thoroughness, completeness, and appropriate standards of description, evidence, and probity rather than flashy claims of superiority—presents an alternative to the existing notions of “excellence” … “Soundness” can be assessed by how it supports socially developed and documentable processes and norms.”1\nThese goals seem to be similar in spirit to the EQUATOR guidelines used in this paper; indeed, this paper could be regarded as an exploratory attempt to implement the “soundness” concept.\nThe variability of institution names is a perennial problem in these types of study: did the authors consider standardising via the Global Research Identifier Database (GRID)? This public domain data set of institutions has been produced by Digital Science as part of their Elements/Altmetrics work and is designed for this type of application.\nScopus does have an open access indicator which is viewable in their web interface: does this value come through the API that was used? If so, it might be worth noting this data as one way to investigate alternative indicators. Also worth mentioning is that at least one university ranking exercise, the SCImago Institutions Rankings, does (from 2019) include a measure that rewards more Open Access publications: find here.\n“It is hard to imagine why most universities continue to support the current ranking schemes given that they may be reducing the positive value universities have on society.” I don’t find this hard to imagine at all, reasons could include: inertia, perceived lack of ability for an individual university to alter the rankings environment, historical autonomy of universities and consequent difficulty of coordinated global action. The institutions at the top of current rankings are probably fairly content with their position and those further down tend to have less influence/power. The paper’s criticism of universities “support” for current ranking schemes can be equally levelled at governments who organise national ranking systems: why should they not evaluate research on the grounds of good practice?\nI would include the term ‘Scopus’ in the Methods section of the Abstract as the data source is a critical aspect of scientometric studies.\nClustering is an appropriate method to use to address the issue of small movements in ranks between years essentially being noise. The clustering definition appears fine although I don’t have experience with Bayesian Clustering so cannot give a fully informed judgement on that specific aspect. However, the key point of the paper does not depend on whether this clustering method is the best or most appropriate. Any broadly equivalent method would be fine. I do not anticipate that the precise numbers/ranks/clusters in this paper would be actually used for ranking institutions in any consequential manner.\nThe key message of the paper is that alternatives to the current ranking systems are feasible. As such it makes a useful contribution to research policy.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-583
https://f1000research.com/articles/8-1203/v1	26 Jul 19	{ "type": "Research Article", "title": "Postoperative pain of patients with necrotic teeth with apical periodontitis following single visit endodontic treatment versus multiple visit endodontic treatment using triple antibiotic paste: a randomized clinical trial", "authors": [ "Safeya AbdurRahman", "Saied M. Abdel Aziz", "Shaimaa I. Gawdat", "Ahmed M. AbdalSamad", "Saied M. Abdel Aziz", "Shaimaa I. Gawdat", "Ahmed M. AbdalSamad" ], "abstract": "Background: A randomized clinical trial was conducted to compare the postoperative pain following endodontic treatment of necrotic teeth with apical periodontitis. Treatments were performed in multiple visits with application of triple antibiotic paste interappointment dressing or single visit without interappointment dressing. Methods: In total 44 participants were assigned randomly into two groups. Group A: multiple visit endodontic treatment with triple antibiotic paste interappointment dressing; group B: single visit endodontic treatment without interappointment dressing. Postoperative pain of participants was assessed after 24, 48, 72 hours and one week using numerical rating scale. Results: No statistically significant difference was found in postoperative pain after 24, 48, 72 hours and one week between the two groups. Conclusion: Triple antibiotic paste as an interappointment dressing in multiple visits endodontic treatment was not proved to reduce the postoperative pain compared to a single visit in patients with necrotic teeth with apical periodontitis who did not have an interappointment dressing. Trial registration: clinicaltrials.gov, NCT02947763. Date: 28th October 2016.", "keywords": [ "Apical periodontitis", "multiple visits", "triple antibiotic paste", "interappointment dressing" ], "content": "Introduction\n\nApical periodontitis (AP) arises primarily by continuous bacterial irritation from the root canals1. AP significantly reduces the endodontic success rate2. Therefore, treatment of AP should aim to completely eliminate the underlying root canal infection3, either by chemomechanical preparation or placement of an interappointment dressing4.\n\nRecent systematic reviews found that multiple visits with calcium hydroxide did not improve postoperative pain and flare-up and radiographic healing compared to single visit endodontic treatment without interappointment dressing5–8. Consequently, the search for a better antimicrobial alternative is required9. Previously, triple antibiotic paste (TAP) has been shown to effectively reduce the bacterial load in necrotic teeth10.\n\nThe aim of the present study was to compare the postoperative pain following endodontic treatment, of necrotic teeth having apical periodontitis, either performed in single visit or in multiple visits with application of triple antibiotic paste interappointment dressing.\n\n\nMethods\n\nThe study was registered on clinical trials.gov and the registration number is NCT02947763.\n\nThe protocol was approved by the Committee of Ethics, Faculty of Dentistry, Cairo University, Egypt (approval no 16562). Participants were asked to sign a printed informed consent that explained the study aim, alternative treatments to endodontic treatment, and the investigator’s instructions.\n\nThe trial design of this study was a parallel, randomized, clinical design with allocation ratio 1:1. This article follows the CONSORT 2010 statement and a copy of the CONSORT checklist can be found in the Data availability section.\n\nThe study began in November 2016 and was completed in February 2018.\n\nPrior data11 indicated that standard deviation of pain score was 20.3. If the true difference in the intervention and control is 20.6, we should study 16 in each group to be able to reject the null hypothesis that the population means of the intervention and control groups are equal with probability (power) 0.8. The Type I error probability associated with this test of this null hypothesis was 0.05. The size of the sample was increased to 22 per group to correct for non-parametric usage and to substitute for any drop-out.\n\nIn total, 44 adults, medically-free with an age range of 16–55 years were selected for the study. All had a necrotic tooth with a periapical lesion confirmed radiographically (minimum size 2 × 2 mm). The participants were enrolled by SA.\n\nExclusion criteria included teeth previously accessed or endodontically treated; vital or necrotic teeth without periapical lesion; patients allergic to metronidazole, ciprofloxacin, or doxycycline or those with significant medical conditions; patients who took analgesic tablets before treatment up to 12 hours previously; and pregnant women.\n\nThe teeth were tested with an electric pulp test (Denjoy DY310 Dental Pulp Tester; Denjoy, Henan, China) to determine pulp sensitivity. Radiographs were taken using a photo stimulable phosphor plate wireless sensor (SOREDEX, DIGORA) to detect periapical lesions.\n\nPreoperative pain was recorded using a numerical rating scale (NRS) where 0 indicates no pain and 10 indicates pain as terrible as it could be12. Pain intensity was categorized into either: none (0); mild (1–3); moderate (4–6); and severe (7–10)13.\n\nLocal anesthesia was administered if needed (Ubistesin™ Articaine HCl 4% & Adrenaline 1:100,000 3M Australia). Isolation of the tooth with rubber dam and preparation of access cavity was performed and the root canal was instrumented by hybrid technique. Coronal shaping was performed with Gates-Glidden drills (MANI, Japan) sizes 4, 3 and 2. Working length was measured using apex locator (J Morita USA) and ascertained using radiograph, where it was set 1 mm away from the radiographic apex. The apical part was instrumented using stainless steel K-files (MANI, Japan); the master apical file size was set 3–4 sizes larger than the initial file. The middle part was instrumented using 3–5 stainless steel H-files.\n\nIrrigation was done using sodium hypochlorite 2.6% (Clorox®, Egypt), using plastic disposable syringe with needle gauge 27, between successive instruments. Lubrication was done using EDTA gel (QMETA, Korea). Final irrigation was done with 5 ml 17% EDTA solution to remove the smear layer (17% EDTA solution, Prevest DenPro Limited, India). The final wash was done using saline. Master cone-fit radiograph was taken to ensure proper length and preparation. The canals were dried with paper points.\n\nAt this step, the participants were divided randomly into two groups with a table of random numbers from 1 to 44 generated by SMA using a freely available computer program with n=22/group. The allocation table was kept with an investigator not involved with participant enrollment (SIG). Numbers from 1 to 44 were written on 44 pieces of paper folded eight-times. Each paper was placed separately in a closed opaque envelope. Each participant was asked to pick one of the envelopes and the participant was assigned to the groups based on the number in the envelope.\n\nGroup A: multiple (two) visit endodontic treatment with triple antibiotic paste interappointment dressing; group B: single visit endodontic treatment without interappointment dressing.\n\nFor group A, triple antibiotic paste (ciprofloxacin, metronidazole, and doxycycline mixed with saline) was prepared and 1 mL of the mixed paste was placed into the canals with a 20-gauge needle of sterile plastic syringe. A sterile cotton pellet was placed, and glass ionomer was placed (Riva Self Cure, Australia). Preparation of the triple antibiotic paste followed the protocol of a previous study14, using ciprofloxacin 250 mg tablets (EPICO, Egypt), metronidazole 500 mg tablets (Aventis, Egypt) and doxycycline 100 mg capsules (Pfizer, Egypt). The powder content of doxycycline capsule was placed in a sterile mortar. In the same mortar, a tablet of metronidazole and a tablet of ciprofloxacin were crushed and all are mixed with saline to a creamy paste. A second appointment was scheduled after at least 7 days. Under rubber dam isolation, the interappointment dressing was removed by H-files and 2.6% sodium hypochlorite and 17% EDTA irrigation followed by saline final wash. Then, obturation was performed using cold lateral condensation technique with resin sealer (ADSEAL, META, Korea). After obturation, glass ionomer was placed to seal the access cavity till final restoration.\n\nFor group B, no interappointment dressing was applied and endodontic treatment was ended in the same visit without placement of interappointment dressing. Obturation and sealing of access cavity were performed as in group A.\n\nPrimary outcomes. Postoperative pain at 24, 48, 72 hours and one week after instrumentation (first visit of Group A and the single visit of Group B); recorded by the participants using NRS in a pain diary.\n\nSecondary outcomes. Incidence of analgesic intake and number of tablets consumed in case of presence of moderate or severe postoperative pain. Participants were instructed to take one tablet of Ibuprofen 400 mg (NOVARITIS, Canada) every 6 hours and to report the number of tablets consumed.\n\nThe operator was blinded until the end of instrumentation until she saw the number of the envelope, then the operator either administered the interappointment dressing in group A or ended the endodontic treatment in a single visit (group B). Blinding of the operator to the end of treatment was difficult as there was only a single operator (SA). The participant did not know whether endodontic treatment was done in multiple or single visit; as another appointment was given to all participants whether to complete the endodontic treatment in Group A or for follow-up in Group B. The data analyst was blinded to the group assignment.\n\nAll the data was collected and tabulated. Statistical analysis was performed by Microsoft Office 2013 (Excel) and statistical package SPSS version 22. The significance level was set at p-value ≤ 0.05. Non-parametric data was summarized as minimum, maximum and median. Chi-squared test was used to compare the incidence of studied parameters and Mann-Whitney test for analysis of severity of pain15.\n\n\nResults\n\nAfter enrollment of 78 patients, only 44 participants were included (Figure 1). The age, gender, and preoperative pain did not differ significantly between the two groups; all participants in the two groups had preoperative no-to-mild pain (Table 1).\n\nMultiple visits group; ♣ Single visit group; NRS = numerical rating scale; SD= standard deviation.\n\nThe data of the postoperative pain are shown in Table 2. There was no statistically significant difference between the two tested groups, either in the intensity nor the incidence of different pain categories. In both groups, there was a significant decrease in the incidence of pain at different follow-up periods (p < 0.05; Table 3).\n\nMultiple visits group; ♣ Single visit group. NRS=numerical rating score.\n\n* Multiple visits group; ♣ Single visit group.\n\nThere was no statistically significant difference between the two tested groups regarding the incidence of analgesic intake and number of analgesics tablets taken by the participants (p > 0.05; Table 4).\n\n*Multiple visits group; ♣ Single visit group.\n\n\nDiscussion\n\nPresence of AP decreases the endodontic success rate, regarding postoperative pain and radiographic healing, by about 10%–15%2,16. Maximum removal of the bacteria and irritants causing AP is essential to achieve better prognosis17. Placement of inter-appointment dressing has been previously recommended to completely disinfect the root canal system4.\n\nTAP has been found to remain active for 30 days18 and shows better antibacterial efficacy than calcium hydroxide in previous in-vitro studies18–28. TAP has been used clinically in case reports and series to treat cases with large periapical lesions, when the use of calcium hydroxide cannot eliminate the symptoms29–32. Moreover, previous randomized clinical trials found that TAP is better than calcium hydroxide, as an intracanal medicament, both clinically and radiographically33–35.\n\nIn the present study, the intensity and the incidence at different pain categories did not differ statistically to a significant level (age, gender and preoperative pain were similarly distributed among both groups). This finding is in accordance to the results of previous studies comparing single visit versus multiple visits with calcium hydroxide36–41. During a previous literature search, the authors found no similar studies comparing single visit versus multiple visits with TAP.\n\nIn our study, the median postoperative pain of multiple visits group (with TAP) at all follow-up periods was 0 and the incidence of moderate and severe pain ranged from 18.2% to 0%. These results are comparable to the results of previous studies. Pai et al.35 found no interappointment flare-up in the group treated with TAP after 1, 2, 3, 7, and 14 days (in diabetic patients). Prasad et al.33 found that the mean of postoperative pain after one week was 0.86, while Bilgi et al.42 found that the mean of postoperative pain after 24 and 48 hours was 0.08.\n\nIn the present study, nearly 95% of the participants were asymptomatic after 72 hours postoperatively. Previous studies found that severe postoperative pain is reduced to a mild pain during this period of time43,44.\n\nWithin the conditions of this study, it could be concluded that postoperative pain was similar after performing endodontic treatment in multiple visits with triple antibiotic paste interappointment dressing or in a single visit.\n\n\nData availability\n\nFigshare: Postoperative Pain after Single versus Multiple Visits Endodontic Treatment of Necrotic Teeth with Apical Periodontitis with Triple Antibiotic Paste: A Randomized Clinical Trial. Dataset demographic data and postoperative pain, https://doi.org/10.6084/m9.figshare.8797592.v145\n\nFigshare: CONSORT checklist, https://doi.org/10.6084/m9.figshare.8797592.v145\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nKatebzadeh N, Sigurdsson A, Trope M: Radiographic evaluation of periapical healing after obturation of infected root canals: an in vivo study. Int Endod J. 2000; 33(1): 60–6. PubMed Abstract \| Publisher Full Text\n\nFriedman S, Abitbol S, Lawrence HP: Treatment outcome in endodontics: the Toronto Study. Phase 1: initial treatment. J Endod. 2003; 29(12): 787–93. PubMed Abstract \| Publisher Full Text\n\nTervit C, Paquette L, Torneck CD, et al.: Proportion of healed teeth with apical periodontitis medicated with two percent chlorhexidine gluconate liquid: a case-series study. J Endod. 2009; 35(9): 1182–5. PubMed Abstract \| Publisher Full Text\n\nLaw A, Messer H: An evidence-based analysis of the antibacterial effectiveness of intracanal medicaments J Endod. 2004; 30(10): 689–94. PubMed Abstract \| Publisher Full Text\n\nAnjaneyulu K, Nivedhitha MS: Influence of calcium hydroxide on the post-treatment pain in Endodontics: A systematic review. J Conserv Dent. 2014; 17(3): 200–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWong AW, Zhang C, Chu CH: A systematic review of nonsurgical single-visit versus multiple-visit endodontic treatment. Clin Cosmet Investig Dent. 2014; 6: 45–56. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFigini L, Lodi G, Gorni F, et al.: Single versus multiple visits for endodontic treatment of permanent teeth: a Cochrane systematic review. J Endod. 2008; 34(9): 1041–7. PubMed Abstract \| Publisher Full Text\n\nSu Y, Wang C, Ye L: Healing rate and post-obturation pain of single- versus multiple-visit endodontic treatment for infected root canals: a systematic review. J Endod. 2011; 37(2): 125–32. PubMed Abstract \| Publisher Full Text\n\nWaltimo T, Trope M, Haapasalo M, et al.: Clinical efficacy of treatment procedures in endodontic infection control and one year follow-up of periapical healing. J Endod. 2005; 31(12): 863–6. PubMed Abstract \| Publisher Full Text\n\nMohammadi Z, Abbott PV: On the local applications of antibiotics and antibiotic-based agents in endodontics and dental traumatology. Int Endod J. 2009; 42(7): 555–67. PubMed Abstract \| Publisher Full Text\n\nPatil AA, Joshi SB, Bhagwat SV, et al.: Incidence of postoperative pain after single visit and two visit root canal therapy: a randomized controlled trial. J Clin Diagn Res. 2016; 10(5): ZC09–12. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWilliamson A, Hoggart B: Pain: a review of three commonly used pain rating scales. J Clin Nurs. 2005; 14(7): 798–804. PubMed Abstract \| Publisher Full Text\n\nJalalzadeh SM, Mamavi A, Shahriari S, et al.: Effect of pretreatment prednisolone on postendodontic pain: a double-blind parallel-randomized clinical trial. J Endod. 2010; 36(6): 978–81. PubMed Abstract \| Publisher Full Text\n\nEstefan BS, El Batouty KM, Nagy MM, et al.: Influence of Age and Apical Diameter on the Success of Endodontic Regeneration Procedures. J Endod. 2016; 42(11): 1620–5. PubMed Abstract \| Publisher Full Text\n\nChan YH: Biostatistics 102: quantitative data--parametric & non-parametric tests. Singapore Med J. 2003; 44(8): 391–6. PubMed Abstract\n\nMarquis VL, Dao T, Farzaneh M, et al.: Treatment outcome in endodontics: the Toronto Study. Phase III: initial treatment. J Endod. 2006; 32(4): 299–306. PubMed Abstract \| Publisher Full Text\n\nNixdorf DR, Moana-Filho EJ, Law AS, et al.: Frequency of persistent tooth pain after root canal therapy: a systematic review and meta-analysis. J Endod. 2010; 36(2): 224–230. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nManiglia-Ferreira C, de Almeida-Gomes F, Pinto MM, et al.: In vitro evaluation of the antimicrobial effects of different intracanal medications in necrotic immature teeth. Eur Arch Paediatr Dent. 2016; 17(4): 251–5. PubMed Abstract \| Publisher Full Text\n\nAdl A, Shojaee NS, Motamedifar M, et al.: A comparison between the antimicrobial effects of triple antibiotic paste and calcium hydroxide against entrococcus faecalis. Iran Endod J. 2012; 7(3): 149–55. PubMed Abstract \| Free Full Text\n\nSabrah AH, Yassen GH, Gregory RL, et al.: Effectiveness of antibiotic medicaments against biofilm formation of enterococcus faecalis and porphyromonas gingivalis. J Endod. 2013; 39(11): 1385–9. PubMed Abstract \| Publisher Full Text\n\nOrdinola-Zapata R, Bramante CM, Minotti PG, et al.: Antimicrobial activity of triantibiotic paste, 2% chlorhexidine gel, and calcium hydroxide on an intraoral-infected dentin biofilm model. J Endod. 2013; 39(1): 115–8. PubMed Abstract \| Publisher Full Text\n\nShokraneh A, Farhad AR, Farhadi N, et al.: Antibacterial effect of triantibiotic mixture versus calcium hydroxide in combination with active agents against Enterococcus faecalis biofilm. Dent Mater J. 2014; 33(6): 733–8. PubMed Abstract \| Publisher Full Text\n\nAdl A, Hamedi S, Sedigh Shams M, et al.: The ability of triple antibiotic paste and calcium hydroxide in disinfection of dentinal tubules. Iran Endod J. 2014; 9(2): 123–6. PubMed Abstract \| Free Full Text\n\nMozayeni MA, Haeri A, Dianat O, et al.: Antimicrobial effects of four intracanal medicaments on enterococcus faecalis: an In vitro study. Iran Endod J. 2014; 9(3): 195–8. PubMed Abstract \| Free Full Text\n\nDevaraj S, Jagannathan N, Neelakantan P: Antibiofilm efficacy of photoactivated curcumin, triple and double antibiotic paste, 2% chlorhexidine and calcium hydroxide against Enterococcus fecalis In vitro. Sci Rep. 2016; 6: 24797. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLakhani AA, Sekhar KS, Gupta P, et al.: Efficacy of Triple Antibiotic Paste, Moxifloxacin, Calcium Hydroxide and 2% Chlorhexidine Gel In Elimination of E. Faecalis: An In vitro Study. J Clin Diagnostic Res. 2017; 11(1): ZC06–ZC09. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMehta S, Verma P, Tikku AP, et al.: Comparative evaluation of antimicrobial efficacy of triple antibiotic paste, calcium hydroxide, and a proton pump inhibitor against resistant root canal pathogens. Eur J Dent. 2017; 11(1): 53–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAbbaszadegan A, Dadolahi S, Gholami A, et al.: Antimicrobial and Cytotoxic Activity of Cinnamomum zeylanicum, Calcium Hydroxide, and Triple Antibiotic Paste as Root Canal Dressing Materials. J Contemp Dent Pr. 2018; 17(2): 105–13. PubMed Abstract \| Publisher Full Text\n\nEr K, Kuştarci A, Ozan U, et al.: Nonsurgical endodontic treatment of dens invaginatus in a mandibular premolar with large periradicular lesion: a case report. J Endod. 2007; 33(3): 322–4. PubMed Abstract \| Publisher Full Text\n\nTaneja S, Kumari M, Parkash H, et al.: Nonsurgical healing of large periradicular lesions using a triple antibiotic paste: A case series. Contemp Clin Dent. 2010; 1(1): 31–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTaneja S, Kumari M: Use of triple antibiotic paste in the treatment of large periradicular lesions. J Investig Clin Dent. 2012; 3(1): 72–6. PubMed Abstract \| Publisher Full Text\n\nDhillon JS, Amita, Saini SK, et al.: Healing of a large periapical lesion using triple antibiotic paste and intracanal aspiration in nonsurgical endodontic retreatment. Indian J Dent. 2014; 5(3): 161–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPrasad LK, Tanwar BS, Kumar KN: Comparison of calcium hydroxide and triple antibiotic paste as intracanal medicament in emergency pain reduction: in vivo study. Int J Care Res. 2016; 4: 244–6. Reference Source\n\nJohns DA, Varughese JM, Thomas K, et al.: Clinical and radiographical evaluation of the healing of large periapical lesions using triple antibiotic paste, photo activated disinfection and calcium hydroxide when used as root canal disinfectant. J Clin Exp Dent. 2014; 6(3): e230–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPai S, Vivekananda Pai AR, Thomas MS, et al.: Effect of calcium hydroxide and triple antibiotic paste as intracanal medicaments on the incidence of inter-appointment flare-up in diabetic patients: An in vivo study. J Conserv Dent. 2014; 17(3): 208–11. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEl Mubarak AH, Abu-bakr NH, Ibrahim YE: Postoperative pain in multiple-visit and single-visit root canal treatment. J Endod. 2010; 36(1): 36–9. PubMed Abstract \| Publisher Full Text\n\nGhoddusi J, Javidi M, Zarrabi MH, et al.: Flare-ups incidence and severity after using calcium hydroxide as intracanal dressing. N Y State Dent J. 2006; 72(4): 24–8. PubMed Abstract\n\nSathorn C, Parashos P, Messer H: The prevalence of postoperative pain and flare-up in single- and multiple-visit endodontic treatment: a systematic review. Int Endod J. 2008; 41(2): 91–99. PubMed Abstract \| Publisher Full Text\n\nInce B, Ercan E, Dalli M, et al.: Incidence of postoperative pain after single- and multi-visit endodontic treatment in teeth with vital and non-vital pulp. Eur J Dent. 2009; 3(4): 273–9. PubMed Abstract \| Free Full Text\n\nAkbar I, Iqbal A, Al-Omiri MK: Flare-up rate in molars with periapical radiolucency in one-visit vs two-visit endodontic treatment. J Contemp Dent Pract. 2013; 14(3): 414–8. PubMed Abstract \| Publisher Full Text\n\nTarale K: Post-operative Pain Analysis between Single Visit and Two Visit Root Canal Treatments using Visual Analogue Scale: An in vivo Study. J Dent Allied Sci. 2013; 2(1): 8–15. Publisher Full Text\n\nBilgi PS, Shah NC, Mehta J: Comparative evaluation of mixture of calcium hydroxide and chlorhexidine, with triple antibiotic paste and combination of calcium hydroxide, chlorhexidine, and lycopene on incidence of interappointment flare-up: an in vivo study. Int J Clin Dent Res. 2017; 1(1): 10–4. Reference Source\n\nGenet JM, Wesselink PR, Thoden van Velzen SK: The incidence of preoperative and postoperative pain in endodontic therapy. Int Endod J. 1986; 19(5): 221–229. PubMed Abstract \| Publisher Full Text\n\nPak JG, White SN: Pain prevalence and severity before, during, and after root canal treatment: a systematic review. J Endod. 2011; 37(4): 429–38. PubMed Abstract \| Publisher Full Text\n\nAbdurrahman S: Postoperative Pain after Single versus Multiple Visits Endodontic Treatment of Necrotic Teeth with Apical Periodontitis with Triple Antibiotic Paste: A Randomized Clinical Trial. figshare. Dataset. 2019. http://www.doi.org/10.6084/m9.figshare.8797592.v1" }	[ { "id": "51774", "date": "12 Aug 2019", "name": "Alireza Adl", "expertise": [ "Reviewer Expertise Endodontics" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article is aimed at comparing the postoperative pain following single or two-visit endodontic treatment of necrotic teeth with apical periodontitis. The subject has been previously investigated when calcium hydroxide was used as the intracanal medicament. The point of this article is using triple antibiotic paste as the inter-appointment medicament. The authors have shown that there is no difference between two groups of patients regarding post-operative pain. Although the subject is interesting, there are a few issues that need to be clarified:\nThe introduction is very short. I believe the authors should explain more about the post-operative pain following endodontic treatment, its mechanisms and incidence. Moreover they also should explain why and how triple antibiotic paste might be effective in reducing the post-operative pain.\n\nIn this study, only necrotic teeth with periapical lesions were included. These teeth are prone to flare up (a combination of pain and swelling). I am interested to know why the authors evaluated only pain and not swelling.\n\nNormally in this kind of study only teeth with a single canal are included. However, in the methods part of this article, it is not clear what types of teeth were selected; single-rooted or multiple-rooted or both types?\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "51777", "date": "13 Aug 2019", "name": "Reham Hassan", "expertise": [ "Reviewer Expertise Endodontics" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe current manuscript links between intra-canal medication, number of visits and the post-operative pain after endodontic treatment. While the results were insignificant, the antibiotic paste presents some disadvantages, like discoloration which was not reported.\n\nOverall, the study was well written and designed. Additionally, it fits in a Randomized Control Trial, which increases the significance of the research.\nIntroduction: Which was well-written and covers the relevant theoretical data of literature.\n\nAim of the study: Here the authors stated the objectives and goals of the study in an appropriate manner.\n\nIn the Methods: The authors gave a detailed description of the employed Allocation concealment and sequence generation was described adequately. The concentration of the creamy antibiotic paste prepared was not mentioned.\n\nThe discussion section fails to explain the results.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "51772", "date": "27 Jan 2020", "name": "A. R. Vivekananda Pai", "expertise": [ "Reviewer Expertise Dr. A. R. Vivekananda Pai: Restorative dental materials", "Restorative and Esthetic dentistry", "and Endodontics. Dr. Htoo Htoo Kyaw Soe: Research Methodology and Biostatistics." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAbstract:\nBackground:\n- Suggested grammatical/formatting corrections:\n1) Page no. 1, Paragraph no. 1, Sentence line no. 4:\nStated as: “dressing or single visit without interappointment dressing.” Suggested correction: dressing or in single visit without an interappointment dressing.\nConclusion:\n- Since the number of visits is not the same/not standardized between intervention and control groups, and is a variable factor between them, the conclusion needs to be revised.\n- Suggested correction for the conclusion:\n1) Page no. 1, Paragraph no. 1, Sentence line no. 1-4:\nStated as: “Triple antibiotic paste as an interappointment dressing in multiple visits endodontic treatment was not proved to reduce the postoperative pain compared to a single visit in patients with necrotic teeth with apical periodontitis who did not have an interappointment dressing.” Suggested correction: There is no significant difference in the postoperative pain when endodontic treatment is carried out in patients for necrotic teeth with apical periodontitis either in multiple visits with an interappointment dressing of triple antibiotic paste or in a single visit without an interappointment dressing.\n\nIntroduction:\n- It is clear and appropriate, but very brief. Therefore, it needs elaboration on the contents related to postoperative pain/flare-up, particularly following single visit vs. multiple visit endodontics, and triple antibiotic paste.\n\nMethods:\n\nTrial design:\n- CONSORT guidelines have been followed and CONSORT checklist for reporting has been mentioned.\n\n- Suggested grammatical/formatting corrections:\n1) Page no. 3, Paragraph no. 2, Sentence line no. 2:\nStated as: “(approval no 16562).” Suggested correction: (approval no. 16562).\n\nOr\n\n(approval number 16562).\n\nSample size calculation:\n- Sample size calculation is clear, but the authors must state the percentage of anticipated drop-out rate.\n- Suggested grammatical/formatting corrections:\n1) Page no. 3, Paragraph no. 1, Sentence line no. 7:\nStated as: “size of the sample was…...” Suggested correction: size of the sample (n) was…...\n\nParticipants:\n- Selection/eligibility criteria have been made explicit.\n- Suggested grammatical/formatting corrections:\n1) Page no. 3, Paragraph no. 1, Sentence line no. 4:\nStated as: “The participants were enrolled by SA.” Suggested correction: The participants were enrolled by author SA.\n\nTreatment procedures:\n- The proportion or ratio of antibiotics used for the TAP should be specified.\n- Suggested grammatical/formatting corrections:\n1) Page no. 3, Paragraph no. 1, Sentence line no. 1:\nStated as: “electric pulp test” Suggested correction: electric pulp tester\n\nRandomization:\n- Details of randomization and allocation concealment have been described.\n- Suggested grammatical/formatting corrections:\n1) Page no. 3, Paragraph no. 1, Sentence line no. 3:\nStated as: “1 to 44 generated by SMA……...” Suggested correction: 1 to 44 generated by author SMA……...\n2) Page no. 3, Paragraph no. 1, Sentence line no. 5:\nStated as: “The allocation table was kept with an investigator not involved with participant enrollment (SIG).” Suggested correction: The allocation table was kept with an investigator not involved with participant enrollment (author SIG).\n3) Page no. 4, Paragraph no. 3, Sentence line no. 5:\nStated as: “glass ionomer was placed……….” Suggested correction: glass ionomer cement was placed……….\n4) Page no. 4, Paragraph no. 3, Sentence line no. 12:\nStated as: “and all are mixed……….” Suggested correction: and all were mixed……….\n5) Page no. 4, Paragraph no. 3, Sentence line no. 19:\nStated as: “glass ionomer was placed……….” Suggested correction: glass ionomer cement was placed……….\n6) Page no. 4, Paragraph no. 4, Sentence line no. 2:\nStated as: “endodontic treatment was ended……….” Suggested correction: endodontic treatment was completed……….\n\nBlinding:\n- Procedure of blinding has been carried out. However, type of blinding is not specified.\n- Type of blinding should be stated.\n- Suggested grammatical/formatting corrections:\n1) Page no. 4, Paragraph no. 1, Sentence line no. 1-2:\nStated as: “The operator was blinded until the end of instrumentation until she saw the number of the envelope,…….” Suggested correction: The operator was blinded until the end of instrumentation and saw the number in the envelope,…….\n2) Page no. 4, Paragraph no. 1, Sentence line no. 3:\nStated as: “or ended the endodontic treatment………” Suggested correction: or completed the endodontic treatment………\n3) Page no. 4, Paragraph no. 1, Sentence line no. 5-7:\nStated as: “Blinding of the operator to the end of treatment was difficult as there was only a single operator (SA). The participant did not know whether endodontic treatment was done in multiple or single visit;……” Suggested correction: Blinding of the operator until the end of treatment was difficult as there was only a single operator (author SA). The participants did not know whether endodontic treatment was done in multiple or in single visit;……\n\nStatistical analysis:\n- Appropriate statistical tests have been carried out.\n- Suggested grammatical/formatting corrections:\n1) Page no. 4, Paragraph no. 1, Sentence line no. 5:\nStated as: “Chi-squared test was……” Suggested correction: Chi-square test was……\n\nResults:\n- All the outcomes measured in the study have been included in the results section.\n- The name of the statistical test/tests applied for each table (i.e. Tables 1, 2, 3, and 4) should be stated as a foot note.\n- The median value of pain intensity, which has been recorded using numerical rating scale (NRS), should be described together with 1st quartile (Q1) and 3rd quartile (Q3).\n- p-value 1 must be reported as 0.999 (Tables 2 and 4).\n- p-values must be reported to 3 decimal places (Tables 1 and 4).\n- In table 4, the total number of tablets taken is shown as 7 under both the groups (Page no. 6, Table no. 4, Row no. 5), whereas the maximum number is 2 and 4 for groups A and B, respectively. This table should be checked to identify whether there is any error in reporting the values.\n\n- Suggested grammatical/formatting corrections:\n1) Page no. 4, Paragraph no. 1, Sentence line no. 3-4:\nStated as: “all participants in the two groups had preoperative no-to-mild pain (Table 1).” Suggested correction: all the participants in both the groups had preoperative no-to-mild pain (Table 1).\n2) Page no. 4, Paragraph no. 2, Sentence line no. 1:\nStated as: “The data of the postoperative pain are shown in Table 2.” Suggested correction: The data of the postoperative pain is shown in Table 2.\n3) Page no. 4, Paragraph no. 2, Sentence line no. 3:\nStated as: “either in the intensity nor the incidence of different pain categories.” Suggested correction: either in the intensity or in the incidence of different pain categories.\n\nOr\n\nneither in the intensity nor in the incidence of different pain categories.\n\n4) Page no. 5, Figure 1, Sentence line no. 10:\nStated as: “Did not receive allocation intervention (n=0).” Suggested correction: Did not receive allocated intervention (n=0).\n5) Page no. 6, Table 3, Row no. 1:\nStated as:\n\n24hrs\n\n48hrs\n\n72hrs\n\n1\n\nweek Suggested correction:\n\n24hrs.\n\n48hrs.\n\n72hrs. 1 week or 1 wk.\n\n6) Page no. 6, Table 3, Row no. 2-3:\nStated as:\n\nGroup A\n\nGroup B\n\nn%\n\nn%\nSuggested correction:\n\nGroup A* [n(%)]\n\nGroup B* [n(%)]\n7) Page no. 6, Table 4, Row no. 1:\nStated as:\n\nGroup A\n\nGroup B\n\nn=22\n\nn=22\nSuggested correction:\n\nGroup A\n\nGroup B\n\n(n=22)\n\n(n=22)\n8) Page no. 6, Table 4, Row no. 3:\nStated as:\n\nn\n\n4\n\n3\n\n%\n\n18.1\n\n13.6\nSuggested correction:\n\n[n(%)]\n\n4 (18.1%)\n\n3 (13.6%)\n\n9) Page no. 6, Table 4, Row no. 4:\nStated as: “No of tablets”\n\nSuggested correction: Number of tablets\n\nOr\n\nNo. of tablets\n\nDiscussion:\n- It needs elaboration with justification on the methodology, particularly emphasizing on the type of tooth selection and the use of hybrid technique of cleaning and shaping.\n- Similarly, it needs more explanation with justification on results.\n- Suggested grammatical/formatting corrections:\n1) Page no. 4, Paragraph no. 2, Sentence line no. 1:\nStated as: “and shows better antibacterial efficacy than calcium hydroxide……...” Suggested correction: and has shown better antibacterial efficacy than calcium hydroxide……...\n\nConclusion:\nStated as: “Within the conditions of this study, it could be concluded that postoperative pain was similar after performing endodontic treatment in multiple visits with triple antibiotic paste interappointment dressing or in a single visit.” Suggested correction: Within the limitations of this study, it could be concluded that there is no significant difference in the postoperative pain when endodontic treatment is carried out in patients for necrotic teeth with apical periodontitis either in multiple visits with an interappointment dressing of triple antibiotic paste or in a single visit without an interappointment dressing.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-1203
https://f1000research.com/articles/8-1189/v1	25 Jul 19	{ "type": "Research Article", "title": "Evaluation of EPAS1 variants for association with bovine congestive heart failure", "authors": [ "Michael P. Heaton", "Adam S. Bassett", "Katherine J. Whitman", "Greta M. Krafsur", "Sang In Lee", "Jaden M. Carlson", "Halden J. Clark", "Helen R. Smith", "Madeline C. Pelster", "Veronica Basnayake", "Dale M. Grotelueschen", "Brian L. Vander Ley", "Adam S. Bassett", "Katherine J. Whitman", "Greta M. Krafsur", "Sang In Lee", "Jaden M. Carlson", "Halden J. Clark", "Helen R. Smith", "Madeline C. Pelster", "Veronica Basnayake", "Dale M. Grotelueschen" ], "abstract": "Background: Bovine congestive heart failure (BCHF) has become increasingly prevalent in feedlot cattle in the Western Great Plains of North America. BCHF is an untreatable complex condition involving pulmonary hypertension that culminates in right ventricular failure and death. A protein variant of hypoxia-inducible factor 2 alpha (HIF2α, encoded by the endothelial PAS domain-containing protein 1 gene, EPAS1) was previously reported to be associated with pulmonary hypertension at altitudes exceeding 2,000 m. Our aim was to evaluate EPAS1 haplotypes for association with BCHF in feedlot cattle raised at moderate altitudes (1,200 m). Methods: Paired samples of clinical cases and unaffected controls were collected at four feedlots in Nebraska and Wyoming. Each pair (n =102) was matched for source, pen, breed type, sex, arrival date, and management conditions. Cases were identified by animal caretakers, euthanized, and diagnosis was confirmed at necropsy. Cases were derived from 30 different ranch operations, with the largest source contributing 32. Animals were tested for eight EPAS1 haplotypes encoding 36 possible different diploid combinations. Results: The common, ancestral EPAS1 haplotype encoding HIF2α with alanine (A) at position 606 and glycine (G) at position 610 was equally frequent in cases and controls (0.67). The EPAS1 variant haplotype reported to be associated with disease (encoding threonine (T) at position 606 and serine (S) at position 610) was not enriched in cases compared with controls (0.21 and 0.25, respectively). Frequencies of other EPAS1 haplotypes (e.g., encoding Q270, L362, or G671) were each less than 0.05 overall. McNemar’s test with 45 discordant pairs showed the linked T606/S610 variant was not associated with BCHF (OR = 0.73, CI95 0.38 -1.4, p-value = 0.37). Conclusions: HIF2α polypeptide variants were not significantly associated with BCHF in feedlot cattle at moderate altitudes. Thus, a wider search is needed to identify genetic risk factors underlying this disease.", "keywords": [ "Heart failure", "cattle", "EPAS1", "HIF2A", "brisket disease", "pulmonary hypertension", "feedlot" ], "content": "Introduction\n\nBrisket disease has been known in the Rocky Mountain Region of Colorado and Utah for more than 100 years as a high altitude disorder characterized by severe ventral edema of the chest tissues1,2. In affected cattle, the reduced partial pressure of oxygen at high altitudes causes pulmonary hypoxia, vascular resistance, arterial remodeling, and pulmonary hypertension. As late as 1963, reports of brisket disease were limited to cattle grazed at altitudes greater than 2,100 m, with a prevalence of 2–10%3. This condition, attributed to ‘cor pulmonale’, eventually causes right ventricular overload and enlargement, ultimately leading to heart failure. In 1976, a similar condition was reported in yearling feedlot cattle maintained at a lower altitude of 1,600 m4. Since the 1970’s, bovine congestive heart failure (BCHF) has become increasingly common in feedlot cattle maintained at the low to moderate altitudes of the North American Plains (800 to 1,600 m)5. However, it is uncertain whether hypobaric hypoxia is the underlying cause of BCHF in these cattle, and evidence suggests that left heart dysfunction may initiate BCHF6. Histopathological assessment of cardiopulmonary tissues obtained from affected cattle fattened at 544-1,420 m revealed significant ventricular fibrosis, abundant cardiac adipose depots, coronary artery injury, and pulmonary venous remodeling6. These features were phenotypically distinct from those of cattle with cor pulmonale at high altitudes. However, other evidence in similarly affected cattle (1,369 m) suggested death occurred prior to the development of advanced obesity7. Thus, disease pathogenesis of BCHF in feedlot cattle maintained at the moderate altitudes remains unclear.\n\nThe impact of BCHF on animal mortality is substantial and appears to be increasing. Mortality in the U.S. and Canada has been estimated at 11 per 10,000 animals entering feedlots, with the rate doubling from 2002 to 20125. Personal communication between authors (MPH and BLVL) and a Nebraska feedlot owner (G. Darnall) in 2018 indicated a BCHF prevalence as high as 7.5% in some single source lots of cattle. Affected cattle were typically bred and managed with the aim of achieving high carcass quality. For some affected producers in the Western Plains, BCHF is their single most costly health-related problem, with losses exceeding $250,000 annually in individual operations, surpassing those from bovine respiratory disease. Consequently, reducing the impact of BCHF is a high priority for the cattle industry.\n\nA potential genetic risk factor for BCHF was reported in 2015 for cattle with high-altitude pulmonary hypertension (1,478 to 2,618 m)8. Angus cattle affected with pulmonary hypertension had a higher frequency of the endothelial PAS domain-containing protein 1 gene (EPAS1) encoding a hypoxia-inducible factor 2 alpha (HIF2α) double variant with threonine (T) at position 606 and serine (S) at 610. This HIF2α T606/S610 variant was proposed to have a dominant gain-of-function activity8. Subsequent whole genome sequencing analysis of EPAS1 in 19 breeds of U.S. cattle identified four additional HIF2α variants encoded by EPAS1 (E270Q, P362L, A671G, and L701F)9. Together, these six amino acid variants comprised eight distinct polypeptide HIF2α sequences. A rooted phylogenetic tree of these HIF2α protein sequences provided a framework for evaluating their potential impact on BCHF in U.S. cattle9. In the present report, our aim was to evaluate EPAS1 haplotypes encoding HIF2α protein variants for association with BCHF in feedlot cattle, raised at moderate altitudes. Here we show that the HIF2α T606/S610 variant encoded by EPAS1 was not associated with BCHF in feedlot cattle maintained at 1,200 m in the geographic region experiencing outbreaks. The results are important for understanding the disease mechanism and for future selection of breeding animals with reduced risk for BCHF.\n\n\nMethods\n\nThe experimental design and procedures used during this research project were approved by the Institutional Animal Care and Use Committee (IACUC) of the University of Nebraska-Lincoln (UNL) Experimental Outline Number 139. The UNL animal care program is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care, registered with the United States Department of Agriculture, and assured by the National Institutes of Health Public Health Service Policy on Humane Care and Use of Laboratory Animals. The university meets these goals through review and approval by the IACUC before projects are initiated for any research and educational activities involving vertebrate animals, to assure compliance with all laws, regulations and rules governing the care and use of animals, and by continuing review and monitoring of approved studies.\n\nNo animals were housed at research facilities or cared for by researchers during this study. All animals were privately owned and located at commercial feedlot operations and managed according to their standard operating procedures (SOP), which includes euthanasia of terminal heart failure cases as recommended by the American Veterinary Medical Association (AVMA). All pen matched unaffected (control) animals were briefly sampled by collecting an ear notch and blood sample and returned to their pens without further involvement in the study. In every instance, efforts were made to ameliorate animal suffering for this incurable, untreatable congesting heart disease.\n\nPaired samples from 102 affected calves and their 102 unaffected matched penmate controls were collected from four private commercial feedlots during a 16-month period spanning January 2017 to April 2018. A sample size of 100 matched pairs was targeted based on the frequency of the EPAS1 T606/S610 variant in Angus cattle (0.22)9. Together with Hardy-Weinberg assumptions, the proportion of discordant pairs having one or two copies of the dominant risk factor was expected to be 0.476 and achieve greater than 80% power to detect an odds ratio of 2.5 using a two-sided McNemar test with a significance level of 0.05 (PASS 2019 Power Analysis and Sample Size Software, version 19.0.2 (NCSS, LLC. Kaysville, Utah, USA).\n\nPotential clinical cases were identified by experienced animal caretakers on horseback (pen riders) and segregated for treatment by moving to a hospital pen. This daily activity is part of the SOP for pen riders, i.e., to identify animals with any potentially serious health problems and move them to a dedicated treatment area. Terminal cases were euthanized by feedlot personnel based on clinical presentation. Euthanasia was accomplished with a well-placed bullet by trained feedlot personnel according to AVMA approved protocols, as part of the feedlot operation’s SOP for terminally ill animals. The case definition included two or more clinical signs specific to BCHF: ventral and intermandibular edema (‘brisket’ and ‘bottle jaw’), jugular vein distention and pulsation, ascites/abdominal swelling, and exophthalmia (‘bug-eye’)10,11. The case definition also included two or more non-specific clinical signs: dyspnea, abducted elbows, depression, drooped ears, intermittent watery orange diarrhea, tachycardia, exercise intolerance, open mouth breathing, and weight loss. Clinical signs increased with disease progression and in some cases, animals died naturally within 24 hours of clinical presentation, while other cases progressed over a period of days to weeks, and thus provided a window of opportunity to collect fresh samples immediately after euthanasia. Researchers were contacted by the feedlot operator when an animal became ill and needed to be euthanized. Researchers then travelled to the feedlot for sample collection. A presumptive diagnosis based on heart morphology and gross lesions was made at necropsy by animal caretakers (cases one to nine) and veterinarians (cases 10 to 102). Carcasses were enrolled in the study only if there was a postmortem presumptive diagnosis of congestive heart failure at necropsy.\n\nControl tissue samples were collected from non-affected penmates matched for source, arrival date, gender, and breed type (based on source, coat color, horned/polled status, ear size and dewlap). The pen sizes ranged from 100 to 300 animals, and control animals were selected based on their willingness to be driven through the gate by riders on horseback. In two instances, the same control animal was inadvertently used as a penmate match for two clinical cases. None of the control animals developed clinical BCHF signs prior to harvest according to feedlot records and pen rider observations.\n\nTissue samples for genomic DNA isolation included V-shaped ear notches and EDTA whole blood collected by venipuncture. Squeeze chute devices at the feedlot facilities were used for sample collection with animals that were able to safely enter and exit the gates. Some affected animals were not able to safely enter the device and thus, their samples were collected immediately after euthanasia. The ear tissue was desiccated with granular NaCl in the field and stored at -20°C upon return to the research facility (four to twelve hours later). The plasma and cellular fractions of EDTA whole blood were separated in the field by centrifugation for 10 minutes at 1,350 x g, placed in liquid nitrogen immediately and stored at -80°C upon return. Hearts, livers, blood, and ear notches from unaffected cattle were collected during federally-inspected beef processing at the U.S. Meat Animal Research Center abattoir from six purebred Angus heifers raised and fattened at 578 m and 15 purebred Angus cows maintained for breeding at 578 m.\n\nUnless otherwise indicated, reagents were molecular-biology grade. DNA from ear notches was extracted by standard procedures12. Briefly, hair follicles were shaved from the tissue and approximately one half of the ear notch was minced and suspended in 2.5 mL of a lysis solution containing 10 mM TrisCl, 400 mM NaCl, 2 mM EDTA, 1% wt/vol sodium dodecyl sulfate, RNase A (250 ug/ml; Sigma-Aldrich, St. Louis, Missouri, USA), pH 8.0. The solution was incubated at 55°C with gentle agitation. After one hour, 1 mg proteinase K was added (Sigma-Aldrich) and the solution was incubated overnight at 55°C with continued agitation. The solution was transferred to a 15 ml tube containing 2 ml of a phase-separation gel (high-vacuum grease, Dow Corning Corporation, Midland, Michigan, USA) and extracted twice with one volume of phenol:chloroform:isoamyl alcohol (25:24:1), and once with one volume of chloroform before precipitation with two volumes of 100% ethanol. The precipitated DNA was washed once in 70% ethanol, briefly air dried, and dissolved in a solution of 10 mM TrisCl, 1 mM EDTA (TE, pH 8.0). A single multiplex matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) assay was used for the six EPAS1 missense SNPs, as previously described9. Assay design and genotyping was performed at Neogen GeneSeek Operations (Lincoln, Nebraska, USA). The iPLEX Gold technology and MassARRAY DNA analysis system with MALDI-TOF MS (Agena Biosciences Inc., San Diego, California) was used according to the manufacturer’s protocol. Briefly, genomic DNA was amplified in microtiter plates using the supplied reagents. After the PCR, excess nucleotides were dephosphorylated by shrimp alkaline phosphatase. This was followed by a single base extension reaction in which a mix of oligonucleotide extension primers was added together with an extension enzyme and mass-modified dideoxynucleotide terminators. The extension primers are designed to anneal directly adjacent to each SNP site and were extended and terminated by a single complementary base. The extension products were desalted and transferred from the microtiter plate onto a chip array, where they crystalize with a pre-spotted MALDI matrix. The chip array was loaded into the mass spectrometer, where the analyte crystals were irradiated by a laser, inducing desorption and ionization. The positively charged molecules accelerate into a flight tube towards a detector. Separation occurs by time-of-flight, which is proportional to the mass of the individual molecules. After each laser pulse, the detector records the relative time of flight for each extension product and the results were displayed on the machine. Genotypes were scored automatically and summary reports were generated.\n\nAs previously reported, a maximum parsimony phylogenetic tree was used to unambiguously phase protein variants encoded by EPAS1 haplotypes9. Haplotype-phased protein variants were unambiguously assigned in individuals that were either: 1) homozygous for all six variant sites, or 2) had exactly one heterozygous variant site. The phylogenetic tree was also important for providing a framework for chi-squared testing. The association of EPAS1 haplotype combinations (diplotypes) with clinical disease was evaluated with two tests. The first was a Pearson’s chi-squared test since it met requirements for appropriate use: nominal categorical data, large sample size (n = 204), and independence of observation assumption (cases were mutually exclusive of controls). There was a small deviation of the assumption that each subject contributes to one and only one cell, because two of the control animals were each inadvertently drawn twice from their pen. None of the animals classified as controls developed clinical disease prior to harvest. A 2 x 4 contingency table was used to test the four diplotypes combinations represented by at least ten animals13. The second evaluation for association of the EPAS1 T610/S610 variant with clinical disease was McNemar’s test for correlated proportions14. This is the most appropriate test for paired nominal data.\n\n\nResults\n\nThe 102 clinical BCHF cases and their penmate controls originated from 30 different sources, with the largest single source contributing 32 matched pairs (Table 1 and S1, see Underlying data). Although breed information was not available for many animals, the pairs were comprised of 100% polled, 93% solid black, and 70% male calves (Table S1, see Underlying data). Most of the 102 calves were from known, well-managed herds that focus on using Angus genetics. Clinical cases of BCHF were documented throughout all stages of calf fattening regimens (Figure 1). The most consistent signs at necropsy included: ventral and intermandibular edema, dilated pulmonary artery and right ventricle, and chronic passive congestion of the liver (Figure 2). The success of identifying BCHF cases was entirely dependent on highly trained, experienced pen riders. The false positive rate for pen rider-identified BCHF cases confirmed at necropsy was zero at sites NE01, NE04, and WY01; and 50% at WY02 (three cases from six suspected cases). Together, samples from the 102 matched case-control pairs represent a resource for evaluating genetic, biochemical, and physiological questions about the mechanism of BCHF in feedlot cattle.\n\naNE and WY sites were located in Nebraska and Wyoming, respectively\n\nThree types of animals identified by feedlot pen riders as end-stage heart failure candidates. Clinical cases were born and raised at 1,000 to 1,200 m prior to feedlot arrival. Their respective heart gross morphologies with enlarged right ventricles and pulmonary arteries are shown below each case. The control heart is from an Angus heifer fattened at 550 m.\n\nPanel A, affected heifer from (pair 77, Table S1) fattened at 1,200 m. Panel B, Gross heart morphology of affected heifer from pair 77 (left) and an unaffected Angus cow maintained for breeding at 578 m (right). The dashed line denotes where each heart was sectioned to show the dorsal half. Abbreviations: RV, right ventricle; LV, left ventricle; PA, pulmonary artery. Panel C, cross-section of the respective livers.\n\nDNA samples from cases and controls were tested for association with known EPAS1 haplotypes according to the phylogenetic framework presented in Figure 3. Haplotype-phased protein variants (diplotypes) were unambiguously assigned for 98.5% of the animals (201/204, Table S1, see Underlying data). The frequency of the common, ancestral EPAS1 haplotype (‘1’) was nearly identical in the cases and controls (0.67). The T606/S610 variant reported to be dominantly associated with high-altitude pulmonary hypertension (haplotype ‘3’) was less frequent in the cases (0.21) than the controls (0.25, Table 2). In addition, there was no statistical difference between four common diplotypes among the cases and controls in a Pearson’s chi-square analysis (Table 3). However, the appropriate statistical test for paired nominal data is McNemar's test. This test with 45 discordant pairs showed the T606/S610 variant was not associated with BCHF (Table 4, OR = 0.73, p-value = 0.37, CI95 0.38 -1.4). With 0.44 discordant pairs having the T606/S610 variant, the sample size of 102 pairs achieved 80% power to detect an odds ratio of 2.5 using a two-sided McNemar test with a significance level of 0.05. Frequencies of three other EPAS1 haplotype variants (Q270, L362, and G671) were less than 0.05 and too rare to analyze with a McNemar test. Thus, an association with EPAS1 haplotype variants was not detected with BCHF in these cattle.\n\nPanel A, map of bovine HIF2α domains in relationship to missense mutations found in cattle: bHLH, basic helix-loop-helix domain; PAS-A and PAS-B, Per-Arnt-Sim domains; ODDD, oxygen-dependent degradation domain; N-TAD, N-terminal transactivation domain; C-TAD, C-terminal transactivation domain. Panel B, HIF2α polypeptide sequences encoded by haplotype variants. The areas of the circles are proportional to the variant frequency in a group of 1,250 cattle from 46 breeds with each node in the tree representing a different polypeptide sequence of HIF2α that varies by one amino acid compared to adjacent nodes. Reproduced from Heaton et al., 20169.\n\naNot detected\n\naDiplotypes not shown had less than ten total observations and thus were omitted from the analysis.\n\nbThe chi-square statistic was 4.3 and the p-value was 0.23.\n\naThe risk factor was defined as having one or two copies of the EPAS1 variant.\n\nbThe ‘(1)’ and ‘(0)’ indicate the presence or absence of the risk factor, respectively.\n\ncThe p-value is the probability of observing this distribution of discordant pairs if there was no association between risk factor and disease.\n\n\nDiscussion\n\nThe present report describes a set of 102 clinical cases of BCHF and their matched controls collected from U.S. feedlots where outbreaks were severe and ongoing. Genetic analyses of EPAS1 variants did not show an association with BCHF. This result was significant because cases were closely matched with unaffected penmate controls from the same source, sex, breed type, arrival date, and management conditions. In some pairs, ear tag information suggested that animals shared the same sire. Case-control studies are a common and efficient means of studying diseases with a low prevalence and the McNemar's test is the appropriate statistical test for use on paired nominal data15. A major advantage of the present design was the ability to increase the frequency of a potential genetic risk factor with relatively few participants (102 cases), while still evaluating more than 10,000 animals per feedlot site for disease. Nevertheless, results presented here suggest it is unlikely that genetic variation at the EPAS1 locus was a significant genetic risk factor for BCHF in these cattle.\n\nThe lack of EPAS1 association with BCHF is inconsistent with that reported in 2015 by Newman et al. for cattle with high-altitude pulmonary hypertension8. However, there were important differences in the cattle types, age, environment, and clinical definitions between the two studies. Animals in the 2015 study were mature, pastured cattle, maintained at 1,478 to 2,618 m, and had pulmonary arterial hypertension, as measured by heart catheterization. In the present study, animals were yearling steers and heifers, raised and fattened at approximately 1,200 m, and had end-stage heart failure. Thus, it is possible that the animals in these respective studies are suffering from similar but different diseases. Another possibility is the cattle described by Newman et al. were affected by a right-sided ventricular heart failure initiated by hypobaric hypoxia, whereas the cattle in the present study had a left-sided ventricular failure, resulting in tissue hypoxia and subsequent BCHF. The latter was reported in 2019 in cattle fattened at 1,200 m suffering end-stage heart failure6. However, it is unknown whether the disease pathogenesis was a distinguishing factor in these studies. One remarkable feature of the BCHF clinical cases in the present study were those displaying signs of the disease at feedlot arrival and soon after. This is consistent with the hypothesis that fattening is not the underlying cause of disease in these cattle.\n\n\nConclusions\n\nProtein variants encoded by EPAS1 haplotypes were not significantly associated with BCHF in fattened cattle at moderate altitudes in the North American Western Plains. Thus, identifying the genetic risk factors underlying this form of the disease may require a wider search.\n\n\nData availability\n\nFigshare: Table S1. Metadata and phased EPAS1 diplotypes for 102 case-control pairs. https://doi.org/10.6084/m9.figshare.8862089.v116\n\nFigshare: Table S2. Summary of 36 possible EPAS1 diplotypes and their frequency in cases and controls. https://doi.org/10.6084/m9.figshare.8862152.v117\n\nFigshare: Table S3. EPAS1 haplotype frequencies in cases and controls. https://doi.org/10.6084/m9.figshare.8862179.v118\n\nFigshare: Table S4. EPAS1 risk factor scoring used in McNemar’s Test with 102 matched pairs. https://doi.org/10.6084/m9.figshare.8862200.v119\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nFunding for this research was provided by the USDA, ARS appropriated projects [5438-32000-033-00D] (MPH) and The University of Nebraska Great Plains Veterinary Educational Center [2162390003] (DMG, BLVL) and the Nebraska Beef Industry Endowment [2662390323001] (BLVL).\n\n\nAcknowledgments\n\nWe thank D. George, D. Ridgway, and S. Schmidt for secretarial and administrative support; S. Johnson for sample collection support; R. Carman, G. Darnall, P. Miller, T. Williams, and their outstanding pen riders for identifying clinical cases, assisting with sample acquisition, and sharing their knowledge. We also thank former USMARC Director, Dr. J. Pollak for leadership and support in the design and execution of this research project. Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the U.S. Department of Agriculture. The USDA is an equal opportunity provider and employer.\n\n\nReferences\n\nGlover GH, Newsom LE: Brisket disease (dropsy of high altitude). Colorado Agricultural Experiment Station. 1915; 204: 3–24. Reference Source\n\nGlover GH, Newsom LE: Further Studies on brisket disease. Journal of Agricultural Records. Washington, DC. 1918; XV(7): 409–13. Reference Source\n\nHecht HH, Kuida H, Lange RL, et al.: Brisket disease. III. Clinical features and hemodynamic observations in altitude-dependent right heart failure of cattle. Am J Med. 1962; 32: 171–83. PubMed Abstract \| Publisher Full Text\n\nJensen R, Pierson RE, Braddy PM, et al.: Brisket disease in yearling feedlot cattle. J Am Vet Med Assoc. 1976; 169(5): 515–7. PubMed Abstract\n\nNeary JM, Booker CW, Wildman BK, et al.: Right-Sided Congestive Heart Failure in North American Feedlot Cattle. J Vet Intern Med. 2016; 30(1): 326–34. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKrafsur GM, Neary JM, Garry F, et al.: Cardiopulmonary remodeling in fattened beef cattle: a naturally occurring large animal model of obesity-associated pulmonary hypertension with left heart disease. Pulm Circ. 2019; 9(1): 2045894018796804. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMoxley RA, Smith DR, Grotelueschen DM, et al.: Investigation of congestive heart failure in beef cattle in a feedyard at a moderate altitude in western Nebraska. J Vet Diagn Invest. 2019; 31(4): 509–522. PubMed Abstract \| Publisher Full Text\n\nNewman JH, Holt TN, Cogan JD, et al.: Increased prevalence of EPAS1 variant in cattle with high-altitude pulmonary hypertension. Nat Commun. 2015; 6: 6863. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHeaton MP, Smith TP, Carnahan JK, et al.: Using diverse U.S. beef cattle genomes to identify missense mutations in EPAS1, a gene associated with pulmonary hypertension [version 2; peer review: 2 approved]. F1000Res. 2016; 5: 2003. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNeary JM, Gould DH, Garry FB, et al.: An investigation into beef calf mortality on five high-altitude ranches that selected sires with low pulmonary arterial pressures for over 20 years. J Vet Diagn Invest. 2013; 25(2): 210–8. PubMed Abstract \| Publisher Full Text\n\nJensen R, Mackey DR: Diseases of feedlot cattle. 3d ed. Philadelphia: Lea & Febiger. 1979; viii: 300.\n\nHeaton MP, Keele JW, Harhay GP, et al.: Prevalence of the prion protein gene E211K variant in U.S. cattle. BMC Vet Res. 2008; 4: 25. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPreacher KJ: Calculation for the chi-square test: An interactive calculation tool for chi-square tests of goodness of fit and independence. 2001. Reference Source\n\nMcNemar Q: Note on the sampling error of the difference between correlated proportions or percentages. Psychometrika. 1947; 12(2): 153–7. PubMed Abstract \| Publisher Full Text\n\nNiven DJ, Berthiaume LR, Fick GH, et al.: Matched case-control studies: a review of reported statistical methodology. Clin Epidemiol. 2012; 4: 99–110. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHeaton M: Table S1. Metadata and phased EPAS1 diplotypes for 102 case-control pairs. 2019. http://www.doi.org/10.6084/m9.figshare.8862089.v1\n\nHeaton M: Table S2. Summary of 36 possible EPAS1 diplotypes and their frequency in cases and controls. 2019. http://www.doi.org/10.6084/m9.figshare.8862152.v1\n\nHeaton M: Table S3. EPAS1 haplotype frequencies in cases and controls. 2019. http://www.doi.org/10.6084/m9.figshare.8862179.v1\n\nHeaton M: Table S4. EPAS1 risk factor scoring used in McNemar’s Test with 102 matched pairs. 2019. http://www.doi.org/10.6084/m9.figshare.8862200.v1" }	[ { "id": "52928", "date": "27 Aug 2019", "name": "Donagh Berry", "expertise": [ "Reviewer Expertise genetic", "genomics" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe report is extremely well written and of high quality; crucially important that \"negative results\" are still published. In the abstract please clarify that the n=102 is 102 pairs - it a bit ambiguous. it would have been nice to genotype the animals on some for of SNP chip to determine the breed composition (you allude to not knowing) and also the pedigree structure (may influence your analysis). Other than for breeding purposes (end of the introduction) the results of such an approach can be used for personalised management (including purchasing). please re-write the 7th line of the second paragraph in the M&M; i can't understand what is being said - maybe there is a missing word or two\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "52931", "date": "03 Sep 2019", "name": "Emily L. Clark", "expertise": [ "Reviewer Expertise Transcriptomics", "ruminants", "functional genomics", "FAANG", "gene expression", "livestock" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe aim of the study described in this manuscript was to evaluate a known genetic risk factor for BCHF in feedlot cattle raised at moderate altitudes by screening of EPAS1 haplotypes in paired cases and controls.\n\nBCHF is a costly health problem for cattle producers in the Western Plains with losses exceeding $250,000 annually and a 11 in 10,000 animal mortality rate. Consequently, reducing the impact of BCHF is a high priority. A potential genetic risk factor for BCHF was reported in 2015 for cattle with high-altitude pulmonary hypertension (1,478 to 2,618 m) but it was unclear at the outset of this study whether this genetic risk was also prevalent in feed-lot cattle at moderate altitudes.\n\nThe authors indicate that \"Affected cattle were typically bred and managed with the aim of achieving high carcass quality.\" Presumably this could indicate there is some pleiotrophic effect of a gene influencing both traits and/or other concomitant traits? Or at least breeding for high carcass quality or other traits might be antagonistic to BCHF prevalence? As such measuring gene expression in lung, liver and heart in affected and unaffected animals in future studies may yield additional information particularly in the pathways that are involved in the disease; I assume the authors might already be working on this.\n\nThe experimental design is clear but defining “pen-matched” and “pen mate” unaffected controls at the first mention in the animal and study design would be helpful to the reader.\n\nWhen the authors indicate that when an animal died naturally this “provided a window of opportunity to collect fresh samples immediately after euthanasia.”, could they provide details of which tissues they collected? Were tissues collected for RNA as well as DNA for genotyping? Presumably samples were also collected for pathology but this isn’t mentioned here.\n\nSimilarly for “Control tissue samples were collected from non-affected penmates”, was this just ear tissue as mentioned in the methodology or other tissues, which could be used for RNA-Seq and gross pathology/histology?\n\n“Hearts, livers, blood, and ear notches from unaffected cattle were collected during federally-inspected beef processing at the U.S. Meat Animal Research Center abattoir from six purebred Angus heifers raised and fattened at 578 m and 15 purebred Angus cows maintained for breeding at 578 m.” It needs to be clear from this sentence that these were for visual and gross morphological comparison with affected BCHF organs, as this only becomes obvious in Figure 1.\n\nSimilarly in figure 2 it is clear that heart and liver were collected to assess gross morphological differences between affected and unaffected animals but this isn’t clear from the animal and study design section.\n\nAs a general comment to summarise the above comments relating to the sample collection could the authors clarify in the animal and study design section which tissues were collected and from which groups of animals in the text, and provide additional details in the form of a table in \"underlying data\" if necessary? This is particularly important because as the authors suggest the samples and information from these cattle are themselves a very useful resource for studying the disease, and knowing exactly which samples have been collected might reduce the chance of duplicated effort in the future.\n\nTechnically I believe the results section of this manuscript is sound. The methodology for genotyping using a custom iPLEX gold genotyping array for the six EPAS1 missense SNPs has been evaluated in previous studies by the authors and to the best of my knowledge the statistical tests used to analyse the results have been interpreted accurately.\n\nIn the discussion the authors mention “In some pairs, ear tag information suggested that animals shared the same sire.” What was the proportion of the total number of animals that shared the same sire? Is it possible to determine how many different sires produced the 102 cases and matched pairs? Were any animals included in the study twins? I am not sure of the twinning rate in these cattle, I assume it might be quite low, but are there ever cases where one twin is affected by BCHF and the other unaffected?\n\nThe lack of an association with the EPAS1 haplotype variants and BCHF, in the feed lot cattle raised at moderate altitude included in this study, is a negative but nevertheless important result, that will be informative to producers, given the importance of BCHF.\n\nThis is a well-designed and executed study that has been presented clearly and informatively by the authors.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "52934", "date": "13 Sep 2019", "name": "Franklyn B. Garry", "expertise": [ "Reviewer Expertise Bovine health and management", "bovine pulmonary hypertension" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an excellent report that investigated whether a specific genetic factor is involved in the development of an increasingly important disease of feedlot animals managed at moderate altitudes. It is scientifically sound and well written. The disease condition is Bovine congestive heart failure, which is clinically similar to altitude-induced pulmonary hypertension with cor pulmonale and right heart failure, commonly known as High Altitude disease or Brisket disease. A previous investigation suggested that a variant of EPAS1 was strongly associated with the development of Brisket disease. This investigation explored whether that gene was associated with BCHF.\n\nThis was a case control study that was well designed and described. The explanation of experimental and statistical methods and results was very clear.\n\nThis is an important area of inquiry because BCHF is an important disease problem with significant impact on feedlot producers, for which there are currently no good preventive, management or treatment practices. Previous work has shown that there are significant differences in the pathologic findings of animals affected by BCHF versus Brisket disease, and this project shows that the genetic basis of the two problems is also likely to be different.\n\nThe authors describe that tissues were collected from some unaffected animals at slaughter, but there was no description of what analyses were performed on these tissues. I am assuming these were used to assess gross morphologic changes in comparison to the tissues of affected animals at necropsy, but this was not described.\n\nThere is a typographical error in the last sentence of the second paragraph of the Methods section, written as 'congesting' rather than 'congestive'.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1189
https://f1000research.com/articles/8-1188/v1	25 Jul 19	{ "type": "Research Article", "title": "Relative prolixity in journals with different citation impact values: an evidence-based scientific writing assessment", "authors": [ "Gyzelle P.V. Nascimento", "Daniel C. Moreira", "Alexis F. Welker", "Gyzelle P.V. Nascimento", "Daniel C. Moreira" ], "abstract": "Background: Scientific writing guidelines recommend that a scientific text should be straightforward, without prolixity, and informative, without obscurity. However, the extent to which researchers follow these recommendations is unknown. Considering that the most cited journals provide more detailed instructions for authors, we aimed to investigate the degree of relative prolixity (i.e., length versus amount of information) among journals with different citation impact scores. Methods: We analyzed journals whose articles follow the classic Introduction, Method, Results, and Discussion structure, written in English and with a CiteScore value ≥ 0.01 classified in the ‘Pharmaceutical Science’ area. Relative prolixity was calculated as the ratio between the number of characters and the number of citations contained in the introductory section of original articles. Additionally, we collected the number of paragraphs and words. Results: The number of characters, words and citations in the Introduction section were significantly higher in the journals with higher CiteScore values. The median number of paragraphs in the Introduction was not affected by the citation impact of the journals. The degree of relative prolixity in the Introduction section of the articles was negatively correlated with the CiteScore values. Conclusions: Articles published in journals with higher CiteScore values have lower degrees of relative prolixity (i.e., shorter texts to transmit a certain amount of information) and obscurity.", "keywords": [ "Academic writing", "Bibliometrics", "Health", "Impact factor", "Scientometrics" ], "content": "Introduction\n\nA scientific text should be straightforward, without prolixity (Matthews & Matthews, 2014), and informative, without obscurity (Annesley, 2010a; Hirst et al., 2015). However, the extent to which researchers follow these recommendations is unknown. Considering that the most cited journals provide more detailed instructions for authors (Gasparyan et al., 2017), we hypothesized that articles published in such journals would be less prolix and more informative than those published in journals with fewer citations. Therefore, the objective of this study was to investigate the degree of relative prolixity (i.e., length versus amount of information) among journals with different CiteScore™.\n\n\nMethods\n\nTo verify the degree of relative prolixity of original articles among journals with different citation impact scores, we chose to analyze journals of a specific area whose publications had their quality recently assessed – ‘Pharmaceutical Science’ area (indexed in the Scopus database) (Bohannon, 2013; Xia et al., 2015). We evaluated 101 journals (from 305) whose articles follow the classic Introduction, Method, Results, and Discussion (IMRAD) structure, written in English and with a CiteScore value ≥ 0.01.\n\nThe degree of relative prolixity was calculated as the ratio between the number of characters, which increase prolixity (Hirst et al., 2015), and the number of citations to any reference, which make the text more informative (Annesley, 2010b; Katz, 2009). Additionally, we collected the number of paragraphs and words. These analyses were done in the Introduction because this is the only section for which scientific writing literature recommends a length limit. To count the number of characters (including spaces), words, paragraphs and citations, the entire body of text of the Introduction section of each article was copied and pasted into Microsoft Office Word 2010 (Microsoft Corporation, Redmond, WA, USA). Each journal was evaluated using the median of the last three published articles in 2018.\n\nJournals were grouped into quartiles of distribution based on their CiteScore values (Fernandez-Llimos, 2018). Differences between groups were assessed using the Kruskal-Wallis test and the correlation between the journals’ CiteScore and the degree of relative prolixity was assessed by the Spearman’s rank correlation test using SPSS software version 21 (IBM, Armonk, NY, USA). For all analyses, p values below 0.05 were considered statistically significant. Data are presented as mean ± standard error of the mean.\n\n\nResults\n\nThe mean CiteScore values of the ‘Pharmaceutical Science’ journals divided into the four quartiles were 0.34, 1.37, 2.40 and 4.14, respectively. The median number of paragraphs in the Introduction of articles did not differ significantly between quartiles and ranged from 3.92 to 4.84 (p = 0.102; Figure 1a). Both the number of characters (including spaces) and citations (21, on average) in the Introduction were significantly higher in the journals with the highest CiteScore (p < 0.001; Figure 1b and c). The number of words in the Introduction gradually increased from the first to the fourth quartile (p < 0.001), whose average number was 442.69, 512.24, 591.00, and 721.60 words. No differences were detected between the first and second, and second and third quartiles (p > 0.05).\n\nMedian number of (a) paragraphs, (b) characters (including spaces), and (c) citations in the Introduction section of journals with different CiteScore values.\n\nThe degree of relative prolixity in the Introduction of the articles presented a negative correlation with the CiteScore value of the journals (p = 0.017; Figure 2).\n\nSpearman’s p = 0.017.\n\n\nDiscussion\n\nThis study showed that the articles published in pharmaceutical science journals with higher values of CiteScore have an Introduction with a lower degree of relative prolixity and more characters, words, and citations. A low degree of relative prolixity matches the consensual recommendation that the Introduction should be as short as possible (Annesley, 2010b; Armağan, 2013; Katz, 2009; Liumbruno et al., 2013; Matthews & Matthews, 2014). One hypothesis to explain our results is that authors who publish in journals with higher citation impact would have better writing skills. In fact, people with less cognitive ability use more words to express information (Saling et al., 2012; Saling et al., 2016). Another hypothesis is that articles that cite more references would simply be more cited, which in fact occurs (consequently increasing CiteScore values) (Fox et al., 2016; Gargouri et al., 2010).\n\nThe higher number of references in articles from journals with higher CiteScore indicates that such articles refer to more experimental findings to support their statements. This agrees with the recommendation that a scientific text should cite the most relevant findings that directly relate to the particular scope of the study (Cals & Kotz, 2013; Seals & Tanaka, 2000; Thrower, 2008), i.e., the “closest information available in the scientific literature” (Katz, 2009). However, the citations may refer to sources that should be avoided, such as books, review articles, and papers whose degree of proximity to the original findings is low (Bavdekar, 2015; Katz, 2009). The average number of words in the Introduction found here (442.69–721.60) was higher than the recommendation of 250–300 words (Bahadoran et al., 2018; Kallestinova, 2011) and disagrees with the consensus that the Introduction should be short (Kallestinova, 2011; Liumbruno et al., 2013; Matthews & Matthews, 2014; Seals & Tanaka, 2000). Furthermore, the number of paragraphs in the Introduction found in the present study (3.92–4.84) was also higher than the recommended number of up to three paragraphs (Annesley, 2010b; Bahadoran et al., 2018; Katz, 2009; Liumbruno et al., 2013; Matthews & Matthews, 2014; Thrower, 2008; Vitse & Poland, 2017), which indicates that many authors are not aware of or ignore this recommendation.\n\nIn conclusion, articles in the pharmaceutical sciences journals with higher CiteScore values show lower degrees of relative prolixity (i.e., shorter texts to transmit a certain amount of information) and obscurity.\n\n\nData availability\n\nFigshare: Nascimento et al. Relative prolixity Dataset 1 CC0. https://doi.org/10.6084/m9.figshare.8872253 (Nascimento et al., 2019).\n\nThis project contains the following underlying data:\n\nNascimento et al. Relative prolixity Dataset 1 CC0.xlsx (Dataset containing the number of characters (including spaces), paragraphs, and citations in the Introduction of the analyzed papers)\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nAnnesley TM: The title says it all. Clin Chem. 2010a; 56(3): 357–360. PubMed Abstract \| Publisher Full Text\n\nAnnesley TM: \"It was a cold and rainy night\": set the scene with a good introduction. Clin Chem. 2010b; 56(5): 708–713. PubMed Abstract \| Publisher Full Text\n\nArmağan A: How to write an introduction section of a scientific article? Turk J Urol. 2013; 39(Suppl 1): 8–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBahadoran Z, Jeddi S, Mirmiran P, et al.: The Principles of Biomedical Scientific Writing: Introduction. Int J Endocrinol Metab. 2018; 16(4): e84795. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBavdekar SB: Writing Introduction: Laying the Foundations of a Research Paper. J Assoc Physicians India. 2015; 63(7): 44–46. PubMed Abstract\n\nBohannon J: Who’s afraid of peer review? Science. 2013; 342(6154): 60–65. PubMed Abstract \| Publisher Full Text\n\nCals JW, Kotz D: Effective writing and publishing scientific papers, part III: introduction. J Clin Epidemiol. 2013; 66(7): 702. PubMed Abstract \| Publisher Full Text\n\nFernandez-Llimos F: Differences and similarities between journal impact factor and CiteScore. Pharm Pract (Granada). 2018; 16(2): 1282. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFox CW, Paine CET, Sauterey B: Citations increase with manuscript length, author number, and references cited in ecology journals. Ecol Evol. 2016; 6(21): 7717–7726. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGargouri Y, Hajjem C, Larivière V, et al.: Self-selected or mandated, open access increases citation impact for higher quality research. PLoS One. 2010; 5(10): e13636. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGasparyan AY, Nurmashev B, Yessirkepov M, et al.: The journal impact factor: Moving toward an alternative and combined scientometric approach. J Korean Med Sci. 2017; 32(2): 173–179. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHirst R, McPherson G, King K: Brevitas and the disabled. IEEE International Professional Communication Conference. 2015. Publisher Full Text\n\nKallestinova ED: How to write your first research paper. Yale J Biol Med. 2011; 84(3): 181–190. PubMed Abstract \| Free Full Text\n\nKatz MJ: From research to manuscript: A guide to scientific writing. Springer, Netherlands. 2009. Publisher Full Text\n\nLiumbruno GM, Velati C, Pasqualetti P, et al.: How to write a scientific manuscript for publication. Blood Transfus. 2013; 11(2): 217–226. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMatthews JR, Matthews RW: Successful scientific writing: a step-by-step guide for the biological and medical sciences. Cambridge University Press. 2014. Publisher Full Text\n\nNascimento GPV, Moreira D, Welker AF: Nascimento et al. Relative prolixity Dataset 1 CC0. figshare. Dataset. 2019. http://www.doi.org/10.6084/m9.figshare.8872253.v1\n\nSaling LL, Willis A, Saling MM: Do the Elderly Get the Message? A Comparative Study of Stories Produced Verbally and as a Text Message. J Psycholinguist Res. 2016; 45(6): 1419–1425. PubMed Abstract \| Publisher Full Text\n\nSaling LL, Laroo N, Saling MM: When more is less: failure to compress discourse with re-telling in normal ageing. Acta Psychol (Amst). 2012; 139(1): 220–224. PubMed Abstract \| Publisher Full Text\n\nSeals DR, Tanaka H: Manuscript peer review: a helpful checklist for students and novice referees. Adv Physiol Educ. 2000; 23(1): S52–58. PubMed Abstract \| Publisher Full Text\n\nThrower PA: Writing a scientific paper: II. Introduction and references. Carbon. 2008; 46(2): 183–184. Publisher Full Text\n\nVitse CL, Poland GA: Writing a scientific paper-A brief guide for new investigators. Vaccine. 2017; 35(5): 722–728. PubMed Abstract \| Publisher Full Text\n\nXia J, Harmon JL, Connolly KG, et al.: Who publishes in “predatory” journals? J Assoc Inf Sci Technol. 2015; 66(7): 1406–1417. Publisher Full Text" }	[ { "id": "54645", "date": "04 Nov 2019", "name": "Asghar Ghasemi", "expertise": [ "Reviewer Expertise I am working on Nitric oxide and diabetes but also have some experience in scientific writing. I teach Scientific writing and have some publications in this field." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper by Nascimento et al. analyzed whether “relative prolixity” of published papers in the field of pharmaceutical science is related to quality of Journals (assessed by CiteScore values). Although, efforts on improving science of writing are acknowledged, in my opinion, the presented paper is subjected to some criticisms as below.\nMajor issues:\nThe index that the authors used to assess prolixity of the published papers seems not to be valid to lead to a legitimate conclusion. The authors did not provide a good background for the index. I could not find any information about “relative prolixity index” in the references that the authors refer to (Hirst et al., 2015; Annesley, 2010; Katz, 2009). There are other scores that authors can use; e.g. “Fog index” (Gunning, 1969[ref-1) which was developed by Robert Gunning to test readability of a paragraph or passage.\n\nThe conclusion of the study is mostly based on the results of a correlation between CiteScore and relative prolixity; the number of journals with high cite score (~10) seems to be low (~10) that affect the correlation coefficient. In addition, correlation per se cannot provide a basis for conclusion. In fact, in this study the authors conclude that the more the references in the introduction, the less the relative prolixity. This is not straightforward as other factors such as topic presented, number of hypotheses being tested, etc. can affect the number of references in the introduction section of a paper. I suggest to consider relevancy of the citations in papers studied.\n\nSome reasoning provided by the authors for the observed association between the index and Journals’ CiteScore values needs revision; the authors stated that “people with less cognitive ability use more words” in their writing. There are more important factors like being a non-native English speaker that influence quality of writing.\nMinor issues:\nCorrelation coefficient and “n” need to be reported in Figure 2.\n\nMultiple comparisons need to be mentioned for ANOVA.\n\nGuides for authors may also affect the results.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [] }, { "id": "56409", "date": "25 Nov 2019", "name": "Farrokh Habibzadeh", "expertise": [ "Reviewer Expertise Medical Journalism" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this manuscript, the authors were supposed to examine the relationship between the \"relative prolixity\" and the CiteScore, an index reflecting the journal quality. They found that articles published in journals with higher CiteScore have lower relative prolixity. There are, however, many major problems with the hypothesis and the methodology used.\n\nMajor Problems:\nThe relative prolixity was defined “as the ratio between the number of characters and the number of citations contained in the introductory section of original articles.” The definition was based on the assumption that increasing the number of characters would increase prolixity, which might be correct, and that citing any reference would make the text more informative, which is correct but NOT relevant to the hypothesis of this study. In fact, although citing any reference would add some information, the number of citations does not necessarily reflect the amount of information conveyed in an article. There is no evidence indicating that the number of citations in the Introduction is associated with the information provided by an article, to the best of my knowledge. Most of the information conveyed in an article is presented in the Results section. Therefore, reasonable surrogates for the amount of information conveyed in an article would be the number of figures in Tables, the number of curves, number of columns in a bar chart, number of points in a scatter plot, number of p values presented (number of hypotheses tested), to name just a few.\n\nAs all the conclusions were made based on an invalid index, the study has neither internal nor external validity.\n\nIn Statistical Methods, the authors should have considered correction of p values reported for multiple comparisons made.\n\nHaving extreme values would increase the absolute value of the correlation coefficient. This should be considered in interpreting the results presented in Fig. 2.\n\nMinor Problems:\nThe Spearman’s rho is not reported.\n\nThere is no definition for “obscurity.” The authors should thus not comment on the obscurity of articles.\n\nIn journalism, we usually refer to journals with the highest CiteScore as Q1, not Q4. In this manuscript, the authors did categorize in an opposite way.\n\nAs Richard Smith, the former Editor of BMJ, has pointed out, many authors do not read the instructions for authors of journals at all, no matter how detailed they are.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [] }, { "id": "56407", "date": "03 Dec 2019", "name": "Kristin Lynn Sainani", "expertise": [ "Reviewer Expertise athletic injuries", "epidemiology", "statistics", "scientific communication" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI commend the authors on addressing a critical problem in the scientific literature: the poor quality of scientific writing. We need more empirical studies that address this important issue, and I hope that the authors will continue to pursue this line of research. I also appreciate that the authors have written the paper in a clear and concise manner. The data were also easy to access and understand. The main finding of interest is that there is a weak inverse correlation between CiteScore (a measure of a journal’s impact) and prolixity (a measure of wordiness). This may be evidence that journals with better writing get more citations, although the evidence for this conclusion is fairly weak. I believe that the paper would be strengthened by considering additional metrics of readability. Specific comments:\nI probably would have chosen the word “wordiness” over “prolixity” in keeping with the theme of encouraging authors to write in a straightforward manner. Prolixity is a more academic term, whereas wordiness is simpler and easier to understand.\n\nI’m not sure why the authors have looked at the number of paragraphs and number of characters as outcomes in themselves. The length of the introduction doesn’t tell us about writing quality or wordiness. A long piece can be well-written and a short piece can be poorly written. Also, different journals have different constraints (such as constraints on word counts) that may affect the length of the introduction. Thus, I don’t think these outcomes are very informative, and need not be highlighted in graphics.\n\nWhy were characters, numbers, and citations analyzed by CiteScore quartile but relative prolixity analyzed treating CiteScore as a continuous variable? In Figure 1, I expected to see a Figure 1d that showed the average prolixity per quartile.\n\nThis paper would be strengthened by considering other metrics of readability beyond just the ratio of characters to citations. There are numerous online tools that allow one to measure readability with validated measures such as the Flesh Reading Ease Scale. See for example, this tool: http://www.checktext.org/. Examining the correlation between CiteScore and readability scores would add to the impact of this paper.\n\nJournals sometimes have limits on the number of references, which would influence the number of citations appearing in the introduction section. The authors should state whether any of the journals examined had such limitations.\n\nThe aim of this study was to gauge the association between a journal’s CiteScore and a measure of the journal’s prolixity. But only 3 samples per journal were taken. Given that prolixity may vary widely from paper to paper within the same journal, I believe that a larger sample size per journal would have strengthened this study. It’s not clear that the last three papers published are going to be representative of all studies in a journal. The authors should comment on how they arrived at the choice of 3 studies per journal.\n\nThe authors should give the magnitude of the correlation coefficient between CiteScore and prolixity. I believe it’s about -.20, which would be considered an extremely weak correlation even though it is statistically significant. The authors should comment on the fact that the correlation is weak. Also, there is one study with high prolixity (>600) that is making it hard to see the pattern of correlation in Figure 2; consider presenting the graph both with and without this point.\n\nBar graphs are not ideal. I would recommend that the authors replace bar graphs with box plots with individual points overlaid, as these are more informative.\n\nThe authors should read and reference this study on text readability in the scientific literature (Plavén-Sigray et al., 20171).\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-1188
https://f1000research.com/articles/7-1349/v1	29 Aug 18	{ "type": "Research Article", "title": "Rice cultivation reduces methane emissions in high-emitting paddies", "authors": [ "Masato Oda", "Chiem Nguyen Huu", "Chiem Nguyen Huu" ], "abstract": "Background: Rice is typically understood to enhance methane emissions from paddy fields. However, rice actually has two separate functions related to methane: i) emission enhancement, such as by providing emission pathways (aerenchyma) and methanogenetic substrates; and ii) emission suppression by providing oxygen pathways, which suppress methanogenesis or enhance methane oxidation. The overall role of rice is thus determined by the balance between its enhancing and suppressing functions. Although existing studies have suggested that rice enhances total methane emissions, we aimed to demonstrate that the balance between rice’s emitting and suppressing functions changes according to overall methane emission levels, which have quite a large range (16‍–500 kg methane ha−1 crop−1). Methods: Using PVC chambers, we compared methane emissions emitted by rice paddy fields with and without rice plants in rice fields in the Mekong Delta, Vietnam. Samples were analyzed by gas chromatograph. Results: We found high overall methane emission levels and our results indicated that rice in fact suppressed methane emissions under these conditions. Emission reductions increased with the growth of rice, up to 60% of emission rate at the maximum tillering stage, then decreased to 20% after the heading stage, and finally recovering back to 60%. Discussion: It is known that methane is emitted by ebullition when the emission level is high, and methane emission reductions in rice-planted fields are thought to be due to oxidation and methanogenesis suppression. However, although many studies have found that the contribution of soil organic matter to methanogenesis is small, our results suggested that methanogenesis depended mainly on soil organic matter accumulated from past crops. The higher the methane emission level, the lower the contribution of rice-providing substrate. Conclusion: As a result, during the growing season, rice enhanced methane emissions in low-emission paddy fields but suppressed methane emissions in high-emission paddy fields.", "keywords": [ "Greenhouse gases", "Mekong Delta", "Methane oxidation", "Methanogenesis inhibition", "Rice paddy", "Triple cropping" ], "content": "Introduction\n\nThe role of rice in methane (CH4) emissions changes according to emission levels. Rice performs three key functions related to CH4 emissions: i) providing a CH4 pathway through a well-developed system of intercellular air spaces (aerenchyma), ii) providing a substrate for methanogenesis, and iii) oxidising CH4 in rhizosphere by supporting O2 counter-transport through aerenchyma system1–6. The level of contribution of these functions varies with the overall emissions, and the total amount of CH4 emitted to the atmosphere is thus a balance between CH4 production and oxidation6. To the best of our knowledge, all studies have concluded that rice enhances overall CH4 emissions from paddy fields; however, these studies have mostly disregarded overall emission levels and their potential impact on the role of rice in enhancing or suppressing CH4 emissions.\n\nCicerone and Shetter measured CH4 emissions with a closed chamber on the water surface of a paddy field, and found that 4 hours after starting measurement, CH4 concentration was 290 ppm over rice plants and only 4 ppm over open water1. Further studies have revealed that CH4 produced under methanogenesis diffuses through the soil, which is oxidised by the surface barrier before reaching the atmosphere2,7. Rice absorbs diffused CH4 from its roots and emits CH4 through aerenchyma3,5,7. Therefore, an established theory has emerged that CH4 is not emitted from the soil without rice. Other recent studies have provided additional evidence that the primary source of CH4 is current-season photosynthates—specifically, root exudates or decaying tissues8–11. This results in CH4 emissions that peak during the late stage of rice growth. Thus, the presence of rice plants has been determined to be the cause of CH4 emissions in paddy fields.\n\nWassmann et al.12 measured CH4 emissions on the water surface of a paddy field amended with organic matter. They found that organic matter incorporation increased total CH4 emission levels from 27–90 to 160–240 kg CH4 ha−1 crop−1, and ebullition increased from 15–23% to 35–62%, respectively12. Since it is known that organic matter incorporation causes CH4 emissions to peak during the early stages of rice growth, when the rice is still small and the aerenchyma is not well developed, the results of Wassmann et al.12 should be closely examined to determine whether ebullition increased with total emissions. According to the 2006 Intergovernmental Panel on Climate Change (IPCC) guidelines, CH4 emissions for 100 days of rice cropping are 130 kg CH4 ha−1 crop−1; however, emissions were observed at almost twice this value by Wassmann et al.12 Average CH4 emissions from rice paddies in Asia without and with organic matter incorporation ranged from 16 to 200 and 250 to 500, respectively13. Although ebullition has been little studied14, ebullition must occur at high emission levels. Furthermore, in the study by Wassmann et al.12, twin CH4 emissions peaks appeared, with an early peak corresponding to the organic matter amendment and a later peak corresponding to rice-originated substrate12. An alternate interpretation of these results is that the twin peaks could reflect the oxidation performance of rice, since CH4 oxidation is known to increase with rice growth, up to the maximum tillering stage, and then decrease15.\n\nWe monitored CH4 emissions in the paddy fields for 5 years (total 15 crops) in triple rice cropping fields in the Mekong Delta, Vietnam. The CH4 emission level was an order of magnitude higher than IPCC standards. These high CH4 emissions suggest that ebullition must have been occurring.\n\n\nResults and discussion\n\nIn the present study, We compared CH4 emissions on the water (soil) surface of paddy fields of with and without rice in paddy fields. Our results showed that rice decreased CH4 emissions by half relative to paddy fields without rice (see Figure 1, Figure 2). Complete, unprocessed data are available on figshare16. There was no marked CH4 emissions peak in the late-stage of the rice field. This suggests that the amount of methanogenesis from rice-providing substrate was relatively small. Note the high emission levels (500–1400 kg CH4 ha−1 crop−1). These findings suggest that total CH4 emissions are reduced by oxidation or methanogenesis inhibition associated with growing the rice plant.\n\nAverage CH4 emissions of rice-planted areas and no-rice-planted areas in a triple cropping rice field in Mekong Delta. Winter–Spring season (2017). The paddy fields did not receive rice straw incorporation. Error bars are s.d. (n = 3).\n\nCumulated CH4 emissions of rice-planted areas and no-rice-planted areas in a triple cropping rice field in Mekong Delta. The number in the legend relates to the plot.\n\nWe also found that the reduction rate of CH4 emissions increased up to 60%, at maximum tillering stage, then decreased to 20% after the heading stage, and then finally recovered to 60% (see Figure 3). The decrease around the heading stage was caused partially by an increase of emissions in rice-planted fields, and mainly by the erratic emissions in the unplanted area (see Figure 1).\n\nDifference in CH4 emissions between rice-planted areas and no-rice-planted areas, calculated by a moving average of five values. The CH4 reduction rate was calculated by (no rice – rice)/no rice.\n\nNo consensus has yet been reached on the extent to which methanotrophs or rice roots attenuate CH4 emissions17. Using the N2 atmosphere technique, CH4 oxidation ratios have been found to be around 40% on average and were relatively stable throughout the rice growing season17. However, genuine CH4 oxidation, as measured using inhibitors, tend to decrease with rice growth, and the reduction rate for total CH4 emissions can reach up to 20%17. Although, most of those studies assume plant-mediated transportation17, in general, our results roughly matched the results using the N2 atmosphere technique. This suggests that the reduction in CH4 emissions was not due to genuine oxidation and was more likely to be due to methanogenesis inhibition by oxygen from aerenchyma, which lasts until harvesting18.\n\nWe found high CH4 emissions in paddy fields that were not planted with rice and did not incorporate rice straw or other organic material. Despite this lower input of methanogenesis substrate, CH4 emission levels were 12 times higher than the IPCC guidelines. The emission levels remained almost stable after reaching maximum. This suggests that methanogenesis mainly depends on soil organic matter that has been accumulated from past rice crops. Prior studies have suggested that the contribution of soil organic matter to methanogenesis is small; however, these studies also indicated that higher emission levels tend to be associated with higher contribution rates of soil organic matter8–11,19. Therefore, our results are consistent with prior studies that assume that emission levels are proportional to the amount of soil organic matter, which can be a methanogenesis substrate. Hotspots of CH4 emissions, which exist widely across tropical Asia, would have huge soil organic matter stock formed by sequential rice cropping under flooded conditions.\n\nTo summarise, most studies of CH4 emissions in paddy fields have been conducted in fields with low overall emission levels. Since the role of rice in CH4 emissions varies according to the overall emission levels, these results cannot be appropriately generalised to rice paddies with high emission levels. The results of our study suggest that rice reduces CH4 emissions in hotspot paddy fields.\n\n\nMethods\n\nExperimental fields were in Tan Loi 2 Hamlet, Thuan Hung village, Thot Not district, Can Tho city, Vietnam. Farmers conducted triple rice cropping by direct seeding and full flooding. This area receives almost 2 months of floods annually from the Mekong River. The flood decomposes rice straw underwater to the extent that it is no obstacle for seeding. Therefore, farmers start the rice cropping by levelling the fields, without incorporating rice straw. We observed CH4 emissions in 18 paddy fields (26 × 17 m each) under several conditions for 5 years from September 2011. A preliminary study was conducted with the rice variety OM501 (suitable for the season) in the Summer-Autumn season of 2016 by the same methods and paddies of the main study (see Figure 4). In the present study, we used three such fields (managed with straw return under flooded conditions) for replication. We conducted the main experiment after the annual floods (4 November 2016–12 February 2017); these fields did not incorporate rice straw.\n\nAverage CH4 emissions of rice-planted areas and no-rice-planted areas in a triple cropping rice field in Mekong Delta. Summer–Autumn season (2016). The paddy fields received incorporated rice straw of the previous crop. Error bars are s.d. (n = 3).\n\nWe compared CH4 emissions in paddy fields with and without rice. We set 2 × 2 m squares of plastic films on each field just before seeding, then carefully removed them with seeds on the films immediately after seeding. In other plots, there was no difference with farmers’ conventional rice-growing procedures. Farmers spread 230 g m−2 (in dry weight) of germinated rice seed (variety Jasmine) on drained wet paddy field surfaces on 5th November. This wet condition was maintained for 7 days, then the field was kept flooded until 89 days after seeding (DAS), and the rice harvested on 100 DAS. The farmers applied fertiliser, which included 76 kg of urea on 12 November, 53 kg each of urea and NPKS (16-16-8-13), diammonium phosphate on 19 November, and 53 kg each of urea and NPKS on 15 December The daily average water level was monitored with a water level logger (HOBO U20; Onset Computer Corporation, Bourne, Massachusetts) at the corner of the field was 2.0 cm (−0.6–6.1 cm) until drained.\n\nWe set an approximately 2 m long and 0.5 m wide ladder from the centre of the shorter bund to allow measurement of CH4 without touching the paddy soil surface. This ladder was on the border of the non-planted area in each plot. We set PVC chamber bases on the paddy field of both sides of the ladder to avoid measurement perturbation. Chambers (60 × 80 cm and 100 cm high, transparent acryl) were set on a watertight chamber base for every measurement. Measurements were taken at 8 a.m. because previous research has indicated that emissions at this time have a high correlation (ca. 90% of average emission) with average daily emissions20. We mixed the air in the chamber with a fan for 5 min after setting the chamber, then sampled the first gas, then sampled the second gas 20 min later. We conducted the measurements once a week throughout the rice growing stage, but every 3 days for 2 weeks after seeding, heading stage, and around draining. The samples were analysed by gas chromatograph (GC-14B, Shimadze, Kyoto). The cumulative CH4 emissions were calculated by linear interpolation.\n\nThis study was conducted with the approval of the farmer.\n\nData were processed using Microsoft Excel 2016.\n\n\nData availability\n\nRaw data of this article is available from figshare: https://doi.org/10.6084/m9.figshare.6916277.v116. Data are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nCicerone RJ, Shetter JD: Sources of atmospheric methane: Measurements in rice paddies and a discussion. J Geophys Res. 1981; 86(C8): 7203–7209. Publisher Full Text\n\nHolzapfel-Pschorn A, Conrad R, Seiler W: Effects of vegetation on the emission of methane from submerged paddy soil. Plant Soil. 1986; 92(2): 223–233. Publisher Full Text\n\nNouchi I, Mariko S, Aoki K: Mechanism of Methane Transport from the Rhizosphere to the Atmosphere through Rice Plants. Plant Physiol. 1990; 94(1): 59–66. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDen van der Gon HA, Van Breemen N: Diffusion-controlled transport of methane from soil to atmosphere as mediated by rice plants. Biogeochemistry. 1993; 21(3): 177–190. Publisher Full Text\n\nWang B, Neue HU, Samonte HP: Role of rice in mediating methane emission. Plant Soil. 1997; 189(1): 107–115. Publisher Full Text\n\nWassmann R, Aulakh MS: The role of rice plants in regulating mechanisms of methane missions. Biol Fertil Soils. 2000; 31(1): 20–29. Publisher Full Text\n\nNouchi I, Mariko S: Mechanism of Methane Transport by Rice Plants. in Biogeochemistry of Global Change. (Springer US), 1993; 336–352. Publisher Full Text\n\nDannenberg S, Conrad R: Effect of rice plants on methane production and rhizospheric metabolism in paddy soil. Biogeochem. 1999; 45(1): 53–71. Publisher Full Text\n\nWatanabe A, Katoh K, Kimura M: Effect of rice straw application on CH4 emission from paddy fields. Soil Sci Plant Nutr. 1993; 39(4): 701–706. Publisher Full Text\n\nTokida T, Adachi M, Cheng W, et al.: Methane and soil CO2 production from current-season photosynthates in a rice paddy exposed to elevated CO2 concentration and soil temperature. Glob Chang Biol. 2011; 17(11): 3327–3337. Publisher Full Text\n\nYuan Q, Pump J, Conrad R: Partitioning of CH4 and CO2 production originating from rice straw, soil and root organic carbon in rice microcosms. PLoS One. 2012; 7(11): e49073. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWassmann R, Neue HU, Alberto MC, et al.: Fluxes and pools of methane in wetland rice soils with varying organic inputs. Environ Monit Assess. 1996; 42(1–2): 163–173. PubMed Abstract \| Publisher Full Text\n\nWassmann R, Neue HU, Lantin RS, et al.: Characterization of methane emissions from rice fields in Asia. I. Comparison among field sites in five countries. Nutr Cycl Agroecosystems. 2000; 58(1–3): 1–12. Publisher Full Text\n\nGreen SM: Ebullition of methane from rice paddies: the importance of furthering understanding. Plant Soil. 2013; 370(1–2): 31–34. Publisher Full Text\n\nDenier van der Gon HAC, Neue HU: Influence of organic matter incorporation on the Methane emission from a wetland rice field. Glob Biogeochem Cy. 1995; 9(1): 11–22. Publisher Full Text\n\nOda M: Rice Cultivation Reduces Methane Emissions in High-Emitting Paddies. figshare. Dataset. 2018. http://www.doi.org/10.6084/m9.figshare.6916277.v1\n\nHayashi K, Tokida T, Kajiura M, et al.: Cropland soil–plant systems control production and consumption of methane and nitrous oxide and their emissions to the atmosphere. Soil Sci Plant Nutr. 2015; 61(1): 2–33. Publisher Full Text\n\nDenier van der Gon HAC, van Breemen N, Neue HU, et al.: Release of entrapped methane from wetland rice fields upon soil drying. Glob Biogeochem Cy. 1996; 10(1): 1–7. Publisher Full Text\n\nWatanabe A, Takeda T, Kimura M: Evaluation of origins of CH4 carbon emitted from rice paddies. J Geophys Res Atmos. 1999; 104(D19): 23623–23629. Publisher Full Text\n\nMinamikawa K, Tokida T, Sudo S, et al.: Guidelines for Measuring CH4 and N2O Emissions from Rice Paddies by a Manually Operated Closed Chamber Method. 2015. Reference Source" }	[ { "id": "39719", "date": "09 May 2019", "name": "Azeem Tariq", "expertise": [ "Reviewer Expertise Greenhouse gas monitoring in agricultural systems." ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors of \"Rice cultivation reduces methane emissions in high-emitting paddies\" present an effort in collecting and measuring the CH4 emissions, in three rice cropping cycles during the 100 days of the field experiment. The overall paper has been written in good English and in a logical manner. The major concern in this manuscript is the methodology and conclusion. The methodology is not clear enough, for example the authors do not make clear the explanation of water treatments in fields with rice and without rice. The authors claimed lower CH4 emissions due to absence of rice plants, but the difference in the first week (Fig. 1) after seeding when there will be hardly any significant role of rice seedlings is not justified.\nFurthermore, the methodology section is much confusing with the use of decomposition of straw and addition of straw (study site section). It is hard to follow the inputs of straw in fields with and without rice. The authors clearly need to modify the methodology section to make the description of treatments and to justify the results.\nFurther, there are no conclusions drawn from the manuscript. The authors forgot to give the conclusion at the end of manuscript.\nIn general, the manuscript is not acceptable in its current condition.\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [ { "c_id": "4659", "date": "22 May 2019", "name": "Masato Oda", "role": "Author Response", "response": "Thank you very much for reviewing our manuscript and giving precious comments.I think your comments can be summarized to three points: “uniformity of conditions”, “the effect of early stage rice plants”, and “the reason of high CH4 emission in the first week”.1) For the first, we would like to apologize for the incomplete explanation and mistype in the treatment subsection that leads readers to misunderstanding for the above points. The sentences and the correct version are as follows: ”We set 2 × 2 m squares of plastic films on each field” → \"We set 2 × 2 m squares of plastic films on a part of each three fields” “In other plots, there was no difference with farmers’ conventional” → ”In other points, there was no difference with farmers’ conventional”These corrections will make clear that the experiment was conducted under completely uniformed conditions of three pair. These corrections are also the answer for the suspicion of uniformity of conditions.2) For the second, we think the effect of rice is large even in the early growth stage of rice because of the high plant density (230 kg ha-1) and the rapid growth of them. We would like to apologize again for mistype of the unit.3) For the third, we think the high CH4 emission in the first week was caused by the erratic character of “ebullition”. You can check there is no difference between \"no rice\" and \"rice\" in the 7th and 9th day’s emission by raw data (https://f1000research.com/articles/7-1349/v1#ref-16).For straws, this is not a treatment but the conventional management of triple cropping rice cultivation in Mekong Delta. They were decomposed in the natural flooding water and not incorporated into the soil (however, straws were incorporated in the preliminary study).Following the editorial team’s recommendation, we are waiting for additional peer review reports before starting on any article revisions." } ] }, { "id": "46457", "date": "28 May 2019", "name": "Kees Jan van Groenigen", "expertise": [ "Reviewer Expertise Agronomy", "biogeochemistry", "greenhouse gas fluxes." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nOda and Chiem present CH4 fluxes from Vietnamese rice paddies with- and without rice plants, and conclude that the presence of rice plants reduces CH4 emissions. Their results are interesting and convincing, and the experimental design seems solid. That said, I have several serious reservations about this paper that need to be addressed before it can be approved.\n\nThe authors need to explain more clearly the relevance of their research. It might be true that the presence of rice plants reduces CH4 emissions, but the sole purpose of paddies is to support rice agriculture. Thus, the statement that rice cultivation reduces methane emissions from paddies seems a bit strange. Does the IPCC provide estimates of emissions from rice paddies without rice plants? If not, how exactly would the results from this study contribute to understanding the contribution of rice agriculture to CH4 emissions? The experimental design does not allow the authors to test their hypothesis as stated in the abstract. The authors aim to demonstrate that the effect of rice plants on methane emissions depends on the background emissions (i.e. without plants). However, the authors only test the effect of rice plants in paddies with high background emissions. The authors did not test the significance of differences between plots with- and without rice plants. The reference list is quite outdated; only 2 of the cited studies were published in the last 5 years. Several relevant recent papers are ignored by the authors. For instance, Jiang et al. 20171 reported that in paddies with high soil C contents, large rice plants reduce CH4 emissions compared to small rice plants. Furthermore, they show the opposite effect in paddies with low soil C contents. Please note that I am a co-author on that paper; I don’t suggest that the authors need to include it, but please consider this as just one example of relevant recent research. The first sentence of the introduction is essentially an hypothesis. Yet, the authors provide no clear support to back up this statement. In the last sentence of the introduction, what does “all studies” refer to? The studies 1-6? Did these studies compare CH4 emissions with and without rice plants? This is not clear. The text is slightly confusing is some places. For instance, in the “background” section in the abstract it’s not clear whether the authors refer to methane emissions with or without rice plants. The same is true for the last paragraph of the introduction. On page 4, second paragraph, the discussion of changes in “reduction rates” is quite confusing. Does a decrease in reduction rates mean means relatively low or high rates in plots with plants? Some basic info about the field site is missing. For instance, MAT and MAP data would be useful, as well as soil C contents. I spotted a few grammatical errors. For instance, in the background section of the abstract, “existing studies” should be “previous studies”. In the discussion section of the abstract “rice-providing” should be “rice-derived”. The first sentence of the results and discussion section should read “In the present study, we compared CH4 emissions on the water (soil) surface of paddy fields with and without rice”. These are just a couple of examples. Perhaps the authors could contact a native speaker to have a quick look at their paper?\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4698", "date": "27 Jun 2019", "name": "Masato Oda", "role": "Author Response", "response": "Thank you for your precious suggestion! We changed the word \"cultivation\" to \"plants\" in the title. We think the word is more suitable to the content. Thank you again. We aim to demonstrate rice plants are suppressing overall methane emission in high CH4 emission paddy fields. We changed the last sentence of the background as follows. \"that the balance...\" to \"that the overall methane emission is decreased by rice plants.\" We expect that this change clarifies some other points that you indicate. In association with, we think the experiment in the low background is not necessary because they have been conducted by plenty of studies. We think it's obvious from Figure 1 and 2; for your information, the p-value by one-sided Welch's t-test for the total emissions is 0.013. Thank you very much for introducing the study to support our conclusion. We referred to that in the last paragraph of the discussion. We have referred to rather old-dated studies because we aim to mention the role of rice in the context of the history of the study for paddy fields CH4 emission. The sentence indicates common sense; therefore, we would like to change as follows: \"Rice is understood to enhance methane emissions from paddy fields as mentioned in IPCC guidelines.\" We would like to change the sentence \", all studies ... these studies\". to \"rice enhances overall CH4 emissions from paddy fields. In addition, previous studies\". We added the formula, and we clarified the sentences in the results section. We think the soil carbon contents is important; however, that is out of range of this study and we don’t have the map. Thank you so much! We had used an English editing service, but that was not enough. We have revised the English." } ] }, { "id": "48479", "date": "14 Jun 2019", "name": "Dung Duc Tran", "expertise": [ "Reviewer Expertise Hydrology and Agricultural Water Management" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors of \"Rice cultivation reduces methane emissions in high-emitting paddies\" present an interesting manuscript with great effort to demonstrate the balance between rice’s emitting and suppressing function changes according to overall methane emission levels. In general, the manuscript is well-written. However, I recommend the authors to do a major revision to the manuscript before being potentially accepted for indexing by the Journal.\nThe Introduction section should add research objectives or in terms of research questions that help readers to promptly understand the paper's content.\n\nIn the Methods section, description and activities are needed in more detail; additionally a map of the case study would be useful.\n\nIn the Results and Discussion section, I expected more discussion that enhance more work on comparing findings with previous studies. In addition, the limitations and future outlooks need to be elaborated upon.\n\nA Conclusion is important to add into the manuscript since it presents the real implication of the research based on the authors' perspective.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [ { "c_id": "4699", "date": "27 Jun 2019", "name": "Masato Oda", "role": "Author Response", "response": "Thank you very much for your precious suggestions. 1. We improved the abstract and the introduction to describe more clearly the aim of this study. 2. We corrected the method section and clarified the details. For the site location map, we think it can be found on the internet nowadays. 3. We added the latest reference which is expected to give readers the deeper understanding of our findings to the result and discussion section. 4. For the limitations and future outlooks, we described them at the end of the discussion section and the newly added conclusion section.See the revised manuscript, please." } ] } ]	1	https://f1000research.com/articles/7-1349
https://f1000research.com/articles/8-387/v1	05 Apr 19	{ "type": "Research Article", "title": "Sex differences in gene expression patterns associated with the APOE4 allele", "authors": [ "Michelle Hsu", "Mehek Dedhia", "Wim E Crusio", "Anna Delprato", "Michelle Hsu", "Mehek Dedhia", "Wim E Crusio" ], "abstract": "Background: The APOE gene encodes apolipoprotein ε (ApoE), a protein that associates with lipids to form lipoproteins that package and traffic cholesterol and lipids through the bloodstream. There are at least three different alleles of the APOE gene: APOE2, APOE3, and APOE4. The APOE4 allele increases an individual's risk for developing late-onset Alzheimer disease (AD) in a dose-dependent manner. Sex differences have been reported for AD susceptibility, age of onset, and symptom progression, with females being more affected than males. Methods: In this study, we use a systems biology approach to examine gene expression patterns in the brains of aged female and male individuals who are positive for the APOE4 allele in order to identify possible sex-related differences that may be relevant to AD. Results: Based on correlation analysis, we identified a large number of genes with an expression pattern similar to that of APOE in APOE4-positive individuals. The number of these genes was much higher in APOE4-positive females than in APOE4-positive males, who in turn had more of such genes than APOE4-negative control groups. Conclusions: Profiling of these genes using Gene Ontology (GO) term classification, pathway enrichment, and differential expression analysis supports the idea of a transcriptional role of APOE with respect to sex differences and AD.", "keywords": [ "APOE4", "Alzheimer's Disease", "Systems Genetics" ], "content": "Introduction\n\nThe APOE gene encodes the apolipoprotein E protein (ApoE), which associates with lipids to form lipoproteins that package and traffic cholesterol and lipids through the bloodstream (de Chaves & Narayanaswami, 2008; Eichner et al., 2002; Puglielli et al., 2003). In the central nervous system, ApoE is primarily produced by astrocytes and carries cholesterol to neurons via ApoE receptors of the low-density lipoprotein receptor (LDLR) family (Getz & Reardon, 2009; Holtzman et al., 2012). There are at least three different alleles of the APOE gene: APOE2, APOE3, and APOE4. In each of these variants, there are two distinct amino acid substitutions at positions 112 and 158 that alter the structure and function of the proteins (Ghebranious et al., 2005; Mahley et al., 2009).\n\nOf the three alleles, APOE3 is the most common and is believed to have a neutral role in Alzheimer disease (AD), whereas APOE2, the least common, is believed to have a protective role. The APOE4 polymorphism, which is found in around 20% of the population but 50% of all patients with AD, increases an individual's risk for developing late-onset AD in a dose-dependent manner (Ashford, 2004; Raber et al., 2004; van der Flier et al., 2006; Zuo et al., 2006).\n\nEvidence suggests that the APOE4 allele may contribute to the risk of developing AD due to an increased amount of amyloid plaques in the brain tissue of affected individuals. In turn, a buildup of amyloid plaques may lead to neuronal degeneration and death, resulting in some of the symptoms associated with AD: memory loss, motor function impairment, dementia, and changes in personality (Barral et al., 2012; Deary et al., 2002; Kim et al., 2009).\n\nHaving the APOE4 mutation does not guarantee that a person will develop AD, indicating that there are other genetic risk factors as well as environment and lifestyle elements that probably contribute to the etiology and progression of the disease.\n\nDifferences have been reported between females and males carrying the APOE4 allele. Women who are positive for APOE4 have a greater risk of developing AD, show accelerated progression of the disease, and have more severe memory and cognitive decline than men with this allele (Altmann et al., 2014; Farrer et al., 1997). In a large scale meta-analysis that considered 58,000 research participants, it was also found that females with the APOE4 allele had a greater risk of developing AD earlier in life than males but that the sex difference was smaller than previously thought (Neu et al., 2017).\n\nIn this study, we examined gene expression patterns associated with aged female and male individuals who are positive for the APOE4 polymorphism in order to identify possible sex-related differences that may be relevant to the onset and progression of AD. Our hypothesis was that gene expression patterns based on correlation analysis with APOE expression in these individuals are different.\n\n\nMethods\n\nThe RNA sequence data used in this study is derived from the brains of an aged population obtained from the Aging, Dementia and Traumatic Brain Injury Study, accessed through the Allen Brain Atlas (http://aging.brain-map.org/). To obtain the data, a gene search specifying APOE was first performed. APOE was then selected and used to extract correlates for each brain region through drop down menus located on the same page. The drop-down menus allowed for the selection of brain region, APOE4-positive or -negative data, and sex. Once the selections were made, the “find correlates” button was used to retrieve the data. The datasets consist of gene expression data together with a clinical diagnosis comparison between males and females. The cohorts used in this study are as follows: APOE4-positive: female 7, males 13; and APOE4-negative: females 31, males 49.\n\nGenes whose expression correlated with that of the APOE gene were collected for the four available brain regions: frontal white matter (FWM) (associated with cognitive function, learning, dementia, and personality changes in AD), hippocampus (HIP), parietal cortex (PCx) (associated with sensory processing, attention, motor function, executive function, and spatial reasoning), and temporal cortex (TCx) (involved in sensory processing as well as declarative and long term memory). Data for each brain region was collected separately for all groups. For information about the RNA-sequence data generation and analysis, see the Quantitative Data Generation white paper in Documentation NA-Seq.\n\nCorrelates to the APOE gene were obtained by querying the database and specifying these parameters: brain region, APOE4 positive and APOE4 negative, and sex. Correlations equal to or greater than \|0.7\| were considered in the analysis. The rationale was to investigate genes with a similar expression pattern as APOE in these different groups to identify genes specific and common to each group, as well as possibly identify genes related to sex differences. Keyword search of the Database for Annotation, Visualization and Integrated Discovery (DAVID) version 6.8 (Huang da et al., 2009) table output was used to identify genes associated with biological processes related to APOE, sex, and AD. Venny 2.0, an online program that compares lists of items, was used to determine the common and unique genes between groups, sex and brain regions.\n\nChi square analyses were used to compare the numbers of gene correlates between the APOE4-positive and APOE4-negative groups and between females and males (Microsoft Office Professional Plus, Excel 2013, Version 15).\n\nThe differential search function of the RNA-Seq page of the Allen Brain Database was used to find genes that show enrichment between females and males, and between males and females for each group and each brain region to identify sex related differences in gene expression patterns. Genes with a 1.5-fold difference expression and greater were evaluated.\n\nDAVID was used to obtain functional information based on Gene Ontology (GO) annotations for the gene correlates. Genes associated with a keyword search term related to APOE, AD, and sex were identified and noted.\n\nKEGG pathway enrichment (using DAVID) was used to further characterize the positive and negative APOE4 gene correlates for each group. Pathways were assessed manually and partitioned based on common themes.\n\n\nResults\n\nWe observed a difference in transcription patterns associated with genes correlating with APOE expression in male and female individuals with the APOE4 allele as compared to APOE4-negative individuals. A difference in transcription patterns between APOE4-positive females and males was also observed.\n\nAs assessed by chi square analyses, the proportions of both positive and negative gene correlates to APOE larger than \|0.7\| were significantly higher in all examined brain regions in both males and females in the APOE4-positive than in the APOE4-negative groups (Extended data Workbooks 1 and 2; Delprato et al., 2019a; Delprato et al., 2019b).\n\nPositive correlates: APOE4-positive vs APOE4-negative groups (all χ2 have 2 df and p<0.001), females: FWM: χ2=4470.1; HIP: χ2=1285.4; PCx: χ2=4050.4; TCx: χ2=3906.60; males: FWM: χ2=574.8; HIP: χ2=602.8; PCx: χ2=187.4; TCx: χ2=318.5.\n\nNegative correlates: APOE4-positive vs APOE4-negative groups (all χ2 have 2 df and p<0.001), females: FWM: χ2=2863.1; HIP: χ2=2654.5; PCx: χ2=2555.1; TCx: χ2=2335.1; males: FWM: χ2=470.8; HIP: χ2=351.4; PCx: χ2=220.3; TCx: χ2=671.1.\n\nIn the APOE4-positive groups we also observed sex differences. Females carrying the APOE4-positive allele had significantly more gene correlates than APOE4-positive males. For all brain regions, differences between female and male groups in the numbers of correlates obtained at or above the cutoff value (r=>\|0.7\|) were assessed with chi square tests (2x2, df=2).\n\nPositive correlates: males vs females (all χ2 have 2 df and p<0.001): FWM: χ2=1926.7; HIP: χ2=237.9; PCx: χ2=2693.4; TCx: χ2=2565.1.\n\nNegative correlates: males vs females (all χ2 have 2 df and p<0.001): FWM: χ2=1051.8; HIP: χ2=1660.9; PCx: χ2=2125.8; TCx: χ2=826.0.\n\nIn the APOE4-negative groups, sex differences were also observed, but they were less strong than in the APOE4 positive groups. For the positive correlates, females had significantly more gene correlates than males. However, for the negative correlates, this was reversed for the FWM, whereas no differences were found for the HIP and TCx.\n\nPositive correlates: males vs females (all χ2 have 2 df and p<0.001 unless indicated otherwise): FWM: χ2=73.6; HIP: χ2=5.3 (0.10>p>0.05); PCx: χ2=6.5 (p<0.05); TCx: χ2=21.8.\n\nNegative correlates: males vs females (all χ2 have 2 df and p<0.001 unless indicated otherwise): FWM: χ2=150.0; HIP: χ2=3.0 (ns); PCx: χ2=84.1; TCx: χ2=0.9 (ns).\n\nThe results indicate that for all brain regions and for both sexes, the number of APOE correlates, and therefore gene expression patterns, are significantly different between the groups carrying the APOE4 allele as compared to APOE4-negative individuals and that expression patterns also differ between females and males of the APOE4-positive groups.\n\nFor the common gene correlates to APOE in APOE4-positive and APOE4-negative individuals, the results for the female and male comparison are as follows: TCx: 47/37 (positive/negative) common genes between females and males, 3263/2532 genes unique to females, 420/577 genes unique to males; HIP 22/16 common genes between females and males, 1164/1984 genes unique to females, 544/323 genes unique to males; FWM 48/22 common gene between females and males, 3465/2335 genes unique to females, 952/818 genes unique to males; PCx 54/42 common gene between females and males, 3432/2605 genes unique to females, 213/175 genes unique to males (Extended data Workbook 3; Delprato et al., 2019c).\n\nResults from a keyword search of GO terms obtained for positive and negative gene correlates for both the APOE4-positive and APOE4-negative male and female groups across the four brain regions show the same trend of enrichment (Figure 1 and Extended data Workbook 4; (Delprato et al., 2019d). Several of the keywords used to analyze the correlates, were associated with a greater number of genes in each brain region and for both positive and negative correlates. The keyword categories are: “immune”, “oxidation”, “inflammation”, “lipid metabolism”, “hormone”, and ”insulin”. Despite the common categories between females and males, there are very few common genes (Extended data Workbook 4; Delprato et al., 2019d).\n\nRepresentative keyword enrichment based on GO Term classification. (A) Females and (B) males, frontal white matter; (C) females and (D) males, hippocampus; (E) females and (F) males, parietal cortex; (G) females and (H) males, temporal cortex. X-axis, keyword categories; Y-axis, frequency of occurrence.\n\nFor the keyword categories, there were no common genes between males and females in the FWM correlates, For the other brain regions common genes between males and females are as follows: HIP positive correlates: “inflammation”, CD4 and CXCL1; negative correlates PPP1CC; PCx negative correlates “Lipid metabolism” PTGES3; “Inflammation” PTGER4; “Oxidation” MTHFD2L; “Immune”, FSD1L, PTGER4; “Hormone” PGRMC2; TCx negative correlates “Inflammation” TBC1D23; “Hormone” GNRH1.\n\nWhether either of the female or male groups had more of any one type of gene is not clear, due to the greater number of gene correlates observed for females. In this respect, the female groups generally had more genes in each of these keyword categories overall. The females also have a substantial number of Alzheimer-related genes associated with each brain region (Extended data Workbook 4; Delprato et al., 2019d).\n\nFemale pathways. To gain further insight into the function of the gene correlates, a pathway analysis was performed for each group (Figure 2 and Extended data Workbook 5; Delprato et al., 2019e). For APOE4-positive females, the strongest pathway themes, when combined, are associated with various types of signaling cascades, such as MAPK (cell cycle, transcription, stress response), Chemokine (immune response), cAMP (2nd messenger signaling), Jak-STAT (immune function, cell division, apoptosis, TNF( immune system function, inflammation, infection response, apoptosis), Toll-like receptor (immune function, innate immunity), neurotrophin (growth factors, protection, development and function of neurons), VEGF (growth factors, vascularization, cancer), hedgehog (cell differentiation and cancer), infection, immune system processes, amino acid metabolism, lipid metabolism, energy metabolism, RNA related processes, cancer, trafficking and recycling organelles (endosome, lysosome, peroxisome).\n\nKEGG pathway categories for female gene correlates. (A) Positive correlates and (B) negative correlates, frontal white matter; (C) positive correlates and (D) negative correlates, hippocampus; (E) positive correlates and (F) negative correlates; parietal cortex; (G) positive correlates and (G) negative correlates, temporal cortex. X-axis, frequency of occurrence; Y-axis, biological function.\n\nThe PCx-positive correlates are associated with the AD, Parkinson, and Huntington pathways. There are 16 genes (ATP5D, ATP6, ATP8, COX1, COX2, COX3, COX6A2, CYTB, NDUFA1, NDUFA13, NDUFA4L2, NDUFB7, NDUFS6, NDUFS7, NDUFS8, UQCR11) in common among these pathways and all are related to energy production, mitochondria related-electron transport and oxidative phosphorylation.\n\nFive genes from the HIP negative correlates associated with the Prion disease pathway: C1QA, C7, C1QB, C6 and C1QC; these are all related to the complement pathway of the innate immune system.\n\nMale pathways. For males, the overall number of pathways associated with the APOE4-positive gene correlates were much fewer than for the females, and in some cases, such as for the PCx-positive correlates and TCx-negative correlates, there were no pathways identified. The greatest number of pathways for males were obtained for the TCx-positive correlates and these were related to fatty acid metabolism, amino acid biosynthesis, cell attachment/cytoskeleton, neuro-related processes such as GABA axon guidance, viral infection, immune function. The HIP positive correlates were related to fatty acid metabolism, amino acid metabolism, Lupus, and viral carcinogenesis. For the HIP-negative correlates two pathways were obtained: alcoholism and RNA transport. For the PCx-negative correlates there were four pathways: spliceosome, RNA degradation, NF-κβ signaling and pertussis. For the other brain regions in males there was one pathway for both positive and negative correlates associated with the FWM: Fatty acid/lipid, carbohydrate metabolism, degradation and glycosphingolipid biosynthesis—lacto and neolacto series (Extended data Workbook 5; Delprato et al., 2019e).\n\nPathways that were exactly the same between female and males were found for the TCx region only, and these are: hsa04510:Focal adhesion, hsa04512:ECM-receptor interaction, hsa05412:Arrhythmogenic right ventricular cardiomyopathy hsa05410:Hypertrophic cardiomyopathy, hsa05414:Dilated cardiomyopathy, hsa04724:Glutamatergic synapse, and hsa00330:Arginine and Proline metabolism.\n\nFor the other brain regions, some of the associated pathways were similar in biological function. Common themes between females and males were related to fatty acid, processes, cardiomyopathy, energy metabolism, amino acid metabolism, cell attachment, ECM (extracellular matrix) receptor interaction, glutamatergic synapse, and pathways related to immune function.\n\nSeveral themes associated with the APOE4-positive females stand out as being distinct from males' pathways and these are related to cancer, signaling, RNA processes, and a myriad of bacterial, viral, and parasitic infectious disease pathways, which include components of endocytosis, intracellular traffic, and immune system function.\n\nGenes expressed higher in females. For all groups and for each brain region in the female and male comparison there is one gene, XIST, associated with X chromosome inactivation, that is highly expressed and female specific (Ji et al., 2015). For the APOE4-positive group and for each brain region there are several genes that have higher expression values for females over males. Seven of these genes are heat shock proteins, which are involved in the cellular stress response, and chaperones, which assist in protein folding and unfolding: BAG3, DNAJB1, CRYAB1, HSPA1A, HSPA1B, HSPA6, HSPB1. The two highest expressed genes aside from XIST are HSPA1A and HSPA1B. The latter two are present in female datasets but not in males. The other top expressing genes in the female and male comparison are RPL9, which is down regulated in severe AD (Kong et al. 2009) and RNU5E-1, small nuclear RNA (Figure 3 and Extended data Workbook 6; Delprato et al., 2019f).\n\n(A) Top differentially expressed genes in the frontal white matter of APOE4-positive females and (B) representative top genes differentially expressed in APOE4-positive males. X-axis, gene symbol; Y-axis, fold difference values.\n\nGenes expressed higher in males. For all groups and brain regions, the same top gene, RPS4Y1, which is on the Y chromosome and encodes for the 40S ribosomal subunit, is highly expressed. RPS4Y1 is an RNA binding protein and it is involved in rRNA processing, translation, and protein targeting. Several other genes are expressed at high levels in males. These include KDM5D (immune system function, oxidation reduction, T-cell antigen processing and presentation, regulation of androgen receptor signaling pathway), USP9Y (ubiquitin/deubiquitination), SNORD3B-2/1 small nuclear RNAs (RNA biogenesis/transport), TXLNGY (syntaxin binding, taxin family), ZFY (transcription regulation), NLGN4Y (lipid metabolism, synapse assembly, neuron cell-cell adhesion), DDX3Y (regulation of gene expression, RNA helicase). With the exception of SNORD3B-2/1, all of the other genes are linked to the Y chromosome and are male specific. Other male genes expressed at lower levels but above the cutoff value have diverse functions, including splicing, immune response, and ubiquitination. Based on GO term annotation, none of the top expressing genes in the male and female comparison are heat shock or chaperone proteins but there is one chaperone protein in males just expressed at a 2.0-fold difference, HBB, which is specific for hemoglobin (Figure 3, Extended data Workbook 6; Delprato et al., 2019f).\n\nWe also examined the differential gene expression pattern of transcription factors associated with APOE4-positive and APOE4-negative individuals and we observed distinct expression of transcription factors between these groups and also between females and males (Extended data Workbook 7; Delprato et al., 2019g).\n\nThere are an increased number of transcription factors expressed in APOE4-positive females as compared with APOE4-positive males. This is also the case for APOE4-positive females as compared to the APOE4-negative female group. For males, the APOE4-negative and APOE4-positive groups have more or less the same set of transcription factors expressed in each brain region with the exception of the FWM in the APOE4-negative group.\n\nTranscription factors overexpressed in females as compared to males for APOE4-positive groups by brain region are as follows. FWM: FOS, JUN, JUNB, S100A8, A100A9, ATF3, HSPA1A, HSPA1B, HMOX1, IL1B, NPAS4, SIK1, ZSCAN31, HIP, FOS, JUNB, S100A12, NME1-NME2, S100A8, S100A9, ATF3, HSPA1A, HSPA1B, HMOX1, NR4A1, VEGFA; PCx; FOS, FOSB, JUNB, NKX6-2, S100A8, S100A9, ATF3, HSPA1A, HSPA1B, ID3 NR4A1, NR4A2, SIK1; TCx: S100A12, S100A9, HSPA1A, HSPA1B.\n\nFor APOE4-negative females, there were no differentially expressed transcription factors observed for the FWM, PCx, and TCx. The exception was for the HIP where HSPA1A and HSPA1B were expressed above the threshold value of 1.5-fold difference.\n\nIn males there are generally the same set of transcription factors for APOE4-positive and APOE4-negative groups associated with the HIP, PCx, and TCx, and three of these are Y chromosome specific: KDM5D, ZFY, and HSFY1.\n\nAn exception was observed for the male APOE4-negative group. Associated with the FWM, there were several additional transcription factors expressed above the threshold value (FOS, CAPN3, ENPP2, GREM1, KDM5D, VEGFA, ZFY and two of these, FOS and VEGFA, also occur in the female APOE4-positive groups (FOS: FWM, HIP, PCx; VEGFA: HIP). Otherwise, the expression of transcription factors observed between females and males is distinct.\n\n\nDiscussion\n\nThe APOE4 allele is the most significant genetic risk factor for late onset AD (Corder et al., 1993; Neu et al., 2017; Raber et al., 2004). It is also associated with aging, atherosclerosis, Lewy Body Dementia, and cardiovascular disease (de Chaves & Narayanaswami 2008; Dickson et al., 2018; Eichner et al., 2002; Riedel et al., 2016). In this study we used a correlation analysis approach to examine gene expression patterns in an aged population consisting of APOE4-positive and APOE4-negative women and men. The goal of this study was to possibly identify sex-related differences in individuals carrying the APOE4 allele. This is because many studies have reported differences in the effects of the APOE4 allele on females particularly, in the context of AD with respect to age of onset and accelerated disease progression associated with dementia and cognitive decline (Farrer et al., 1997; Podcasy & Epperson, 2016; Riedel et al., 2016; Vest & Pike, 2013). Our hypothesis was that gene expression patterns in APOE4-positive and APOE4-negative individuals would differ. We found a large number of genes with the same expression pattern as APOE in APOE4-positive, but not APOE4-negative individuals. This trend was larger in APOE4-positive females as compared with APOE4-positive males and with APOE4-negative control groups.\n\nWe did not anticipate that such a large number of gene correlates observed within the significance range would differ between the APOE4-positive and APOE4-negative groups and also between APOE4-positive females and males, even though the expectation had been that there would be differences related to gene function. Despite the differences in the overall numbers of correlates between the groups, the same major categories of genes were observed for both positive and negative correlates as well as for all of the four brain regions considered in this study. The categories of genes relate to immune processes, oxidation, inflammation, lipid metabolism, and hormones.\n\nThe results are largely consistent with what is known about APOE function. It is interesting, but in retrospect not surprising, that the same gene categories would emerge for all groups despite the substantial differences in the numbers of gene correlates which may be associated with the different biological processes affected by the APOE4 allele.\n\nThe results of the pathway analysis were somewhat similar, but some differences were also observed. A large number of pathways were obtained for the female gene correlates, but many fewer for the males, which is not surprising given the larger number of gene correlates for females. Those that could be compared between the two sexes show that there are common themes such as immune function, lipid metabolism, and inflammation. Interestingly, many of the female pathways were related to microbial infection, major signalling circuits, and amino acid metabolism. Several other pathways were concerned with RNA, cancer, and cardiovascular processes. These are very diverse but all relatable to APOE function and biological programs.\n\nIn the evaluation of genes expressed at higher levels in APOE4-positive females than in males, several genes were observed exclusively in females; these were either chaperones that function in protein folding and unfolding or heat shock proteins that play a protective role in cellular stress. The role of these proteins in AD has been investigated (Campanella et al., 2018; Hamos et al., 1991; Marino Gammazza et al., 2016). In contrast, the genes expressed higher in males than in females were Y-chromosome specific. Distinct patterns of transcription factor expression were also observed between the groups. The transcription factors that are expressed in APOE4-positive females are involved in the regulation of a multitude of diverse processes. Several of these have established roles in the regulation of immune system response, inflammation, oxidative stress, aging, and estrogen signaling, which may be of particular relevance to this study (Extended data Workbook 7; Delprato et al., 2019g).\n\nOthers have reported differences in gene expression patterns associated with the APOE4 allele (Huang et al., 2017; Lin et al., 2018; Theendakara et al., 2018). The APOE4 allele has also been shown to alter the transcriptional profile of 857 genes with similar expression patterns to APOE in induced neurons and glia (Lin et al., 2018). The major pathways identified for these genes are consistent with what we observe: lipid processes, immune system, inflammation, energy metabolism, and transcription. There is a growing body of evidence indicating that APOE has transcription factor activity. Most of the studies in support of this were done in vitro and in mouse models. Further testing in humans is imperative and may provide a clearer understanding of the relationship between APOE, aging, and AD.\n\nThe results from our study are consistent with known APOE-related processes and provide further support for the regulatory role of APOE as a transcription factor. That our findings are consistent with other reports that have used different models to study the relationship of APOE with AD and aging support the approach that we have taken (Huang et al., 2017; Lin et al., 2018; Theendakara et al., 2018). We provide further relevance and insight for the differential transcription-related effects of the APOE4 allele in APOE4-positive males and females, which may be related to the overall heightened immune response and increased risk for AD observed in females (Klein, 2012; Taneja, 2018). The identification of specific genes, gene families, and pathways that are affected by the APOE4 allele will help to dissect its complex role in diverse but interrelated physiological processes. It would be interesting to compare the results from our study with APOE4 expression patterns in a younger population to see if these differences between APOE4-positive and -negative groups and between females and males are already present and detectable early on or are age dependent.\n\n\nData availability\n\nUnderlying RNA sequence data used in this study was obtained from the Aging, Dementia and Traumatic Brain Injury Study, accessed through the Allen Brain Atlas (http://aging.brain-map.org/).\n\nFigshare: Workbook 1. Positive and negative gene correlates for APOE4+ and APOE4- females. Frontal white matter, Hippocampus, Parietal cortex, and Temporal cortex. https://doi.org/10.6084/m9.figshare.7849175 (Delprato et al., 2019a)\n\nWorkbook 2. Positive and negative gene correlates for APOE4+ and APOE4- males. Frontal white matter, Hippocampus, Parietal cortex, and Temporal cortex. https://doi.org/10.6084/m9.figshare.7849190 (Delprato et al., 2019b)\n\nWorkbook 3. Common and unique genes for APOE4+ and APOE4- males and females. Frontal white matter, Hippocampus, Parietal cortex, Temporal cortex https://doi.org/10.6084/m9.figshare.7849196 (Delprato et al., 2019c)\n\nWorkbook 4. Keywords for gene correlates APOE4+ and APOE4- females and males. Frontal white matter, Hippocampus, Parietal cortex, Temporal cortex https://doi.org/10.6084/m9.figshare.7849214 (Delprato et al., 2019d)\n\nWorkbook 5. KEGG Pathways for positive and negative correlates associated with APOE4+ and females and males. Frontal white matter, Hippocampus, Parietal cortex, Temporal cortex https://doi.org/10.6084/m9.figshare.7849229 (Delprato et al., 2019e)\n\nWorkbook 6. Differential gene expression in APOE4+ and APOE4- females and males. Frontal white matter, Hippocampus, Parietal cortex, Temporal cortex https://doi.org/10.6084/m9.figshare.7849238 (Delprato et al., 2019f)\n\nWorkbook 7. Differentially expressed transcription factors in females and males. GO Biological Process. Frontal white matter, Hippocampus, Parietal cortex, Temporal cortex https://doi.org/10.6084/m9.figshare.7849253 (Delprato et al., 2019g)\n\nExtended data are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nAltmann A, Tian L, Henderson VW, et al.: Sex modifies the APOE-related risk of developing Alzheimer disease. Ann Neurol. 2014; 75(4): 563–573. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAshford JW: APOE genotype effects on Alzheimer's disease onset and epidemiology. J Mol Neurosci. 2004; 23(3): 157–165. PubMed Abstract \| Publisher Full Text\n\nBarral S, Bird T, Goate A, et al.: Genotype patterns at PICALM, CR1, BIN1, CLU, and APOE genes are associated with episodic memory. Neurology. 2012; 78(19): 1464–1471. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCampanella C, Pace A, Caruso Bavisotto C, et al.: Heat Shock Proteins in Alzheimer's Disease: Role and Targeting. Int J Mol Sci. 2018; 19(9): pii: E2603. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCorder EH, Saunders AM, Strittmatter WJ, et al.: Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer's disease in late onset families. Science. 1993; 261(5123): 921–923. PubMed Abstract \| Publisher Full Text\n\nde Chaves EP, Narayanaswami V: Apolipoprotein E and cholesterol in aging and disease in the brain. Future Lipidol. 2008; 3(5): 505–530. PubMed Abstract \| Free Full Text\n\nDeary IJ, Whiteman MC, Pattie A, et al.: Cognitive change and the APOE epsilon 4 allele. Nature. 2002; 418(6901): 932. PubMed Abstract \| Publisher Full Text\n\nDelprato A, Crusio W, Dedhia M, et al.: Workbook 1. Positive and negative gene correlates for APOE4+ and APOE4- females. Frontal white matter, Hippocampus, Parietal cortex, and Temporal cortex. figshare. Fileset. 2019a.http://www.doi.org/10.6084/m9.figshare.7849175.v1\n\nDelprato A, Crusio W, Dedhia M, et al.: Workbook 2. Positive and negative gene correlates for APOE4+ and APOE4- males. Frontal white matter, Hippocampus, Parietal cortex, and Temporal cortex. figshare. Fileset. 2019b. http://www.doi.org/10.6084/m9.figshare.7849190.v1\n\nDelprato A, Crusio W, Dedhia M, et al.: Workbook 3. Common and unique genes for APOE4+ and APOE4- males and females. Frontal white matter, Hippocampus, Parietal cortex, Temporal cortex. figshare. Fileset. 2019c. http://www.doi.org/10.6084/m9.figshare.7849196.v1\n\nDelprato A, Crusio W, Dedhia M, et al.: Workbook 4. Keywords for gene correlates APOE4+ and APOE4- females and males. Frontal white matter, Hippocampus, Parietal cortex, Temporal cortex. figshare. Fileset. 2019d. http://www.doi.org/10.6084/m9.figshare.7849214.v1\n\nDelprato A, Crusio W, Dedhia M, et al.: Workbook 5. KEGG Pathways for positive and negative correlates associated with APOE4+ females and males. Frontal white matter, Hippocampus, Parietal cortex, Temporal cortex. figshare. Fileset. 2019e. http://www.doi.org/10.6084/m9.figshare.7849229.v2\n\nDelprato A, Crusio W, Dedhia M, et al.: Workbook 6. Differential gene expression in APOE4+ and APOE4- females and males. Frontal white matter, Hippocampus, Parietal cortex, Temporal cortex. figshare. Fileset. 2019f. http://www.doi.org/10.6084/m9.figshare.7849238.v1\n\nDelprato A, Crusio W, Dedhia M, et al.: Workbook 7. Differentially expressed transcription factors in females and males.GO Biological Process. Frontal white matter, Hippocampus, Parietal cortex, Temporal cortex. figshare. Fileset. 2019g. http://www.doi.org/10.6084/m9.figshare.7849253.v2\n\nDickson DW, Heckman MG, Murray ME, et al.: APOE ε4 is associated with severity of Lewy body pathology independent of Alzheimer pathology. Neurology. 2018; 91(12): e1182–e1195. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEichner JE, Dunn ST, Perveen G, et al.: Apolipoprotein E polymorphism and cardiovascular disease: a HuGE review. Am J Epidemiol. 2002; 155(6): 487–495. PubMed Abstract \| Publisher Full Text\n\nFarrer LA, Cupples LA, Haines JL, et al.: Effects of age, sex, and ethnicity on the association between apolipoprotein E genotype and Alzheimer disease. A meta-analysis. APOE and Alzheimer Disease Meta Analysis Consortium. JAMA. 1997; 278(16): 1349–1356. PubMed Abstract \| Publisher Full Text\n\nGetz GS, Reardon CA: Apoprotein E as a lipid transport and signaling protein in the blood, liver, and artery wall. J Lipid Res. 2009; 50 Suppl: S156–61. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGhebranious N, Ivacic L, Mallum J, et al.: Detection of ApoE E2, E3 and E4 alleles using MALDI-TOF mass spectrometry and the homogeneous mass-extend technology. Nucleic Acids Res. 2005; 33(17): e149. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHamos JE, Oblas B, Pulaski-Salo D, et al.: Expression of heat shock proteins in Alzheimer's disease. Neurology. 1991; 41(3): 345–350. PubMed Abstract \| Publisher Full Text\n\nHoltzman DM, Herz J, Bu G: Apolipoprotein E and apolipoprotein E receptors: normal biology and roles in Alzheimer disease. Cold Spring Harb Perspect Med. 2012; 2(3): a006312. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHuang da W, Sherman BT, Lempicki RA: Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc. 2009; 4(1): 44–57. PubMed Abstract \| Publisher Full Text\n\nHuang YA, Zhou B, Wernig M, et al.: ApoE2, ApoE3, and ApoE4 Differentially Stimulate APP Transcription and Aβ Secretion. Cell. 2017; 168(3): 427–441.e21. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJi B, Higa KK, Kelsoe JR, et al.: Over-expression of XIST, the Master Gene for X Chromosome Inactivation, in Females With Major Affective Disorders. EBioMedicine. 2015; 2(8): 909–918. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKim J, Basak JM, Holtzman DM: The role of apolipoprotein E in Alzheimer's disease. Neuron. 2009; 63(3): 287–303. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKlein SL: Sex influences immune responses to viruses, and efficacy of prophylaxis and treatments for viral diseases. BioEssays. 2012; 34(12): 1050–1059. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKong W, Mou X, Liu Q, et al.: Independent component analysis of Alzheimer's DNA microarray gene expression data. Mol Neurodegener. 2009; 4: 5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLin YT, Seo J, Gao F, et al.: APOE4 Causes Widespread Molecular and Cellular Alterations Associated with Alzheimer's Disease Phenotypes in Human iPSC-Derived Brain Cell Types. Neuron. 2018; 98(6): 1141–1154.e7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMahley RW, Weisgraber KH, Huang Y: Apolipoprotein E: structure determines function, from atherosclerosis to Alzheimer's disease to AIDS. J Lipid Res. 2009; 50 Suppl: S183–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMarino Gammazza A, Bavisotto CC, Barone R, et al.: Alzheimer's Disease and Molecular Chaperones: Current Knowledge and the Future of Chaperonotherapy. Curr Pharm Des. 2016; 22(26): 4040–4049. PubMed Abstract \| Publisher Full Text\n\nNeu SC, Pa J, Kukull W, et al.: Apolipoprotein E Genotype and Sex Risk Factors for Alzheimer Disease: A Meta-analysis. JAMA Neurol. 2017; 74(10): 1178–1189. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPodcasy JL, Epperson CN: Considering sex and gender in Alzheimer disease and other dementias. Dialogues Clin Neurosci. 2016; 18(4): 437–446. PubMed Abstract \| Free Full Text\n\nPuglielli L, Tanzi RE, Kovacs DM: Alzheimer's disease: the cholesterol connection. Nat Neurosci. 2003; 6(4): 345–351. PubMed Abstract \| Publisher Full Text\n\nRaber J, Huang Y, Ashford JW: ApoE genotype accounts for the vast majority of AD risk and AD pathology. Neurobiol Aging. 2004; 25(5): 641–650. PubMed Abstract \| Publisher Full Text\n\nRiedel BC, Thompson PM, Brinton RD: Age, APOE and sex: Triad of risk of Alzheimer's disease. J Steroid Biochem Mol Biol. 2016; 160: 134–147. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTaneja V: Sex Hormones Determine Immune Response. Front Immunol.. 2018; 9: 1931. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTheendakara V, Peters-Libeu CA, Bredesen DE, et al.: Transcriptional Effects of ApoE4: Relevance to Alzheimer's Disease. Mol Neurobiol. 2018; 55(6): 5243–5254. PubMed Abstract \| Publisher Full Text\n\nvan der Flier WM, Schoonenboom SN, Pijnenburg YA, et al.: The effect of APOE genotype on clinical phenotype in Alzheimer disease. Neurology. 2006; 67(3): 526–527. PubMed Abstract \| Publisher Full Text\n\nVest RS, Pike CJ: Gender, sex steroid hormones, and Alzheimer’s disease. Horm Behav. 2013; 63(2): 301–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZuo L, van Dyck CH, Luo X, et al.: Variation at APOE and STH loci and Alzheimer's disease. Behav Brain Funct. 2006; 2: 13. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "47507", "date": "30 Apr 2019", "name": "J.Wesson Ashford", "expertise": [ "Reviewer Expertise Psychiatry", "neuroscience" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article by Hsu et al. is an interesting and worthwhile article.\n\nMy 3 problems with such articles are:\nGrouping genotypes as e4 positive or negative – the e4/4 group is as different from the e3/4 group as the e3/4 group is from the e3/3 group, and the e2/4 is very similar to e3/3 – so this is poor data analysis. There is an exponential increase of dementia/Alzheimer rate with age, doubling every 5 years, which is rarely taken into account. Since women live on the average 5 years longer than men, this totally ignored discrepancy can fully account for the male/female difference in Alzheimer incidence, and should definitely be discussed in such studies. Only precise age adjustments with an exponential term can be correctly interpreted. Most such data sets use 5-year cohorts, which is why this issue is wrongly considered. To really resolve Alzheimer’s disease, more precision of simple epidemiological data is needed. A major issue is Leptin, which has a level twice as high in women as men. Leptin is related to energy metabolism and learning, both important issues in AD.\n\nFor yet another perspective, see Smith et al., 20191.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [] }, { "id": "47508", "date": "17 May 2019", "name": "Yu-Wen Alvin Huang", "expertise": [ "Reviewer Expertise Neuroscience", "molecular biology", "stem cell biology", "neurodegeneration", "Alzheimer's disease", "aging" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe allelic variant of apolipoprotein E (ApoE) gene, APOE-e4, is the strongest genetic risk factor for Alzheimer’s disease (AD), with a clearly greater female than male predisposition (reviewed in Belloy et al., 20191). Despite the prominence and significance of this APOE-by-sex interaction, the underlying mechanism remains to be elucidated. To tackle this critical issue in a comprehensive fashion, Hsu et al. mined the rich data sets of The Aging, Dementia and Traumatic Brain Injury Study, which comprise structured information including transcriptomics, clinical and pathological findings based on 107 aged brains (http://aging.brain-map.org/). Their functional analyses of genes correlating to APOE expression support our current understanding of ApoE's pleiotropy in multiple brain functions and, of note, reveal the female-specific pathways presumably driven by a sex-dependent, transcriptional role of ApoE. While their results are informative and very interesting to the AD research community, we would like to suggest the authors to clarify or elaborate a few points, which, in our view, will complement this solid manuscript and better inspire future studies for this underappreciated matter of gender-associated AD risk:\n\nThe fundamental approach here is to analyze APOE gene correlates – genes whose expression correlated with that of APOE – by a built-in function of the aforementioned database online platform, followed by Gene Ontology characterization and pathway enrichment (DAVID). By this approach, the correlation was first computed with normalization to APOE transcript level change (=1.0). We recommend that the APOE expression pattern to be examined in different groups, including ApoE genotypes, genders, ages and clinico-pathological stages. To further exploit this valuable database for better insight on the reported results, the information regarding the disease severity - CERAD score and Braak stage – can be used to investigate whether the striking ApoE4-positive female pathways operate at an early stage (more likely to be proactive), at a mid-to-late stage (more likely to be reactive) or along the disease progression (like a disease marker). The authors listed a number of APOE-correlated genes that are most differentially expressed in comparison of genders (Figure 3), and most of them represent a well-known cellular function that is important but not conventionally considered to contribute to AD pathogenesis. In the same context of gender-specific ApoE4 effect, do authors mind providing their observation on the expression pattern of other top AD genes (e.g. APP, PSEN1/2, TREM2, etc.) as well as some of risk genes identified in recent GWAS (e.g. Jensen et al., 20192)? The results may not be of significant difference but will be of great benefit for understanding how ApoE4 contributes to AD pathogenesis in women.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-387
https://f1000research.com/articles/8-1169/v1	23 Jul 19	{ "type": "Research Article", "title": "In silico study of ginsenoside analogues as possible BACE1 inhibitors involved in Alzheimer's disease", "authors": [ "Neyder Contreras-Puentes", "Jairo Mercado-Camargo", "Antistio Alvíz-Amador", "Jairo Mercado-Camargo", "Antistio Alvíz-Amador" ], "abstract": "Background: Neurodegenerative disorders such as Alzheimer's disease show an alarming prevalence in the population, with around 50 million affected individuals worldwide, and is associated with the development of dementia, mainly affecting the elderly population. Currently, the treatments used are based on slowing the progression of the clinical symptoms of Alzheimer’s; however, no specific treatment has been found that induces reversion of the disease. Natural products may induce a wide range of palliative effects, such as ginseng (Panax ginseng), which produces secondary metabolites called ginsenosides, which have multiple therapeutic applications, including for neurodegenerative diseases such as Alzheimer’s. Methods: A virtual screening was carried out, using the CHEMBL database to find analogs of ginsenosides based on the ginsenoside Rg1 (CHEMBL501637). Next, the molecules were optimized and their format modified. The structure of BACE1 was downloaded from the Protein Data Bank database (ID: 1FKN) and prepared for the development of molecular docking using the PyRx software. A database of the ligands was created and molecular docking experiments were carried out, obtaining affinity energy values in Kcal/mol. Results: Of the 27 analogues studied, it was found that the molecules CHEMBL451292, CHEMBL510371 and CHEMBL503302 showed considerable affinity with BACE1 when compared with the reference molecules (verubecestat and donepezil). These affinity energy values for CHEMBL451292, CHEMBL510371 and CHEMBL503302 were -9.6, -8.1 and -7.6 Kcal/mol, respectively. Likewise, the binding site of the ligands to BACE1 were identified, with the main interactions being van der Waals and hydrogen bonding. Conclusions: Three ginsenoside analogues showed a similar binding energy with BACE1 compared to the reference drugs. The residues involved in the inhibitory activity of BACE1 and the type of predominant interactions were identified, which agreed with previous reports.", "keywords": [ "BACE1", "ginsenoside", "molecular docking", "inhibitor", "Alzheimer´s disease" ], "content": "Introduction\n\nAlzheimer's disease is a neurodegenerative condition causing changes in behavioral patterns and progressive cognitive deterioration, leading to dementia and diminished orientation and recognition abilities1,2. It has been estimated that more than 45 million individuals have been diagnosed with dementia worldwide and projections for prevalence in 2050 indicate that around 130 million people will develop these types of manifestations3,4.\n\nAt present, Alzheimer's disease leads to irreversible repercussions for the diagnosed population, which constitutes a problem at the public health level. Current treatments are based on the functioning of neurotransmitters such as acetylcholine, specifically as acetylcholinesterase inhibitors, preventing the degradation of this endogenous substance and favoring the stimulation of associated neurotransmitter response mechanisms5,6. However, these treatments are lacking and are inefficient at significantly prolonging quality of life for patients.\n\nTherefore, we have studied the properties of products of natural origin such as ginseng (Panax ginseng), specifically major secondary metabolites such as ginsenosides7, which are extracted from different plant components. These molecules generally contain steroid rings and have varied biological responses, such as anti-inflammatory, anti-carcinogenic and cardiovascular effects and, recently, activities at the level of nervous regulation, with potential effects on epilepsy, Parkinson’s and Alzheimer’s8–10.\n\nTherefore, in this study, promising molecules from structural variants of ginsenosides with potential capacity of bound to proteins involved in the development of Alzheimer's disease such as β-secretase (BACE-1) were identified using interaction studies, identification of affinities, determination of binding-free energies and in silico methodologies based on molecular docking.\n\n\nMethods\n\nA search for ginsenoside Rg1 (CHEMBL501637) analogues was performed using the CHEMBL database11. The name of the structure ginsenoside Rg1 (CHEMBL501637) was used to search the database. Subsequently, a filter was applied to the analogues to identify compounds with 80–90% similarity. The database provided information about the physical representation, activity charts, calculated properties and solubility of these compounds. Previous theoretical studies with ginsenosides have reported finding few structures upon screening. A preliminary set database of 50 analogue structures with the highest similarities was developed. In order to evaluate this first set of molecules, over half of the structures (27) were randomly selected using a random number generator (accession numbers for these molecules are provided as Underlying data and their structures as Extended data). Two drugs were selected as references, which were searched for using the ZINC Database12 and PubChem, then selected and saved in MOL2 format. The selected structures were Donepezil (ZINC accession number 597013), due to its use in pharmacological therapy for mild and moderate Alzheimer’s disease. Likewise, verubecestat (PubChem accession number 51352361) was developed as a candidate molecule for the selective inhibition of BACE1.\n\nBIOVIA Discovery Studio software version 4.513 was used to correct the structures through the conversion of the format of the structure, addition of hydrogen atoms, neutralization of the charged groups, generation of ionization states, clean geometry and, finally, the conversion of the output file into its final MOL 2 format. Subsequently, the representative crystallographic structure of BACE1 (Accession number 1FKN, resolution of 1.9 Å) was obtained from Protein Data Bank (PDB)14. For the preparation of the receptor, the protein obtained in PDB format was prepared through the addition of hydrogen atoms, the elimination of solvent molecules (water) and the allocation of partial charges using the software packages UCSF Chimera version 1.1315 and BIOVIA Discovery Studio version 4.516.\n\nThe molecular docking was performed with AutoDock Vina 4.2.117, using a PyRx working interface18. Prior to molecular docking, a virtual screening of 27 ginsenoside analogues was carried out to determine the molecules with the best structural affinity and candidates for a potential inhibitory effect of BACE1, whereby analogues were minimized by universal force field and conjugate gradients using a default algorithm based on Open Babel tools. BACE1 and ginsenosides analogue molecules were represented in a grid space of x = 25 Å, y = 25 Å, z = 25 Å A using a universal force field. Subsequently, each docking simulation used a single ginsenoside analogue molecule and the software selected the top eight conformations with the largest change in free energy of binding, which were classified according to the energy level of each conformation and according to root-mean-square-deviation value obtained. The result of each docked molecule was determined according to the docking energy of the receptor-ligand complex. The top conformations of the structures with BACE1 were obtained in pdbqt format, using PyMOL software version 2.3.219 for its visualization and conversion into PDB format. BIOVIA Discovery Studio version 4.5 was used to determine interaction forces and amino acids.\n\n\nResults\n\nDocking interactions between the ginsenoside analogues with the highest binding affinity with BACE1 are shown in Figure 1A–E. The six analogues show binding energies of ≥7,5 Kcal/mol, affinity values that are comparative to those of the reference molecules. Equally, reference drugs, inhibitors of BACE1 and of acetylcholinesterase (AChE), were used in the study are shown in Figure 2A and 2B. Different types of interactions were observed between analogues and the residues of the pocket formed in the BACE1 enzyme, such as van der Waals, H-bonding (conventional and carbon), π-sigma, alkyl and π-alkyl. Likewise, common amino acids involved in binding with ginsenosides analogues such as L30, D32, S35, Y71, T72, Q73, I110, D228, T231 and T232 were identified. The binding affinity of selected ginsenosides, the BACE1 inhibitor and the acetylcholinesterase inhibitor with BACE1 are shown in Table 1, where it can be seen that the compound CHEMBL451292, CHEMBL510371, CHEMBL503302, CHEMBL201643, CHEMBL448647 and CHEMBL508182, showed the highest energy affinity, with values of -9.6, 8.1, 7.6, 7.5, 7.5 and 7.5 Kcal/mol, respectively. On the other hand, the reference drugs showed energies of binding 9.8 Kcal/mol for verubecestat and 8.3 Kcal/mol for donepezil. The hydrophobic residues Q12, I110, G11, T231, T232, I118, L30, G34, S35, D32, and N233 were found to strengthen the interactions between BACE1 and ginsenosides analogues through van der Waals interactions. It was also found that hydrogen bonding interactions between the active site of the BACE1 enzyme and ginsenoside analogues involved residues D32, Q73, N233, G264, R307, R235, T72, F322, K321, T231 and T329.\n\nMolecular interactions of ginsenoside analogues and β-secretase 1 (BACE1 A. CHEMBL451292 analogue binding to BACE1. B. CHEMBL510371 analogue binding to BACE1. C. CHEMBL503302 analogue binding to BACE1. D. CHEMBL201643 analogue binding to BACE1. E. CHEMBL448647 analogue binding to BACE1. F. CHEMBL508182 analogue binding to BACE1.\n\nA. Verubecestat (inhibitor) binding to BACE1. B. Donezepil binding to BACE1.\n\nBACE1, β-secretase 1; AChE, Acetylcholinesterase.\n\n\nDiscussion\n\nIt has been reported that ginsenosides demonstrate a variety of biological activities and due to this, have been progressively included as alternative in cardiovascular, antioxidant and anticancer therapies. Their recently discovered activities at the neuronal level establish a reason for the study of these metabolites in regards to neurodegenerative alterations such as Alzheimer's disease4,20, which are not fully understood at the molecular level. For example, the study of ginsenosides as anti-Alzheimer’s agents, specifically as potential BACE1 inhibitors21. Therefore, some authors have proposed experimental studies that indicate the potential role of these compounds as promising elements in the inhibition of BACE1 and as future molecules for the treatment of this pathology. Therefore, as an alternative to a molecular study and before the empirical approaches, in silico studies have been implemented to reveal the capacity binding and affinity of these metabolites and BACE1, studying the interaction and affinity energy of 27 randomly selected ginsenoside analogues.\n\nThe molecular docking experiments in this study revealed that the higher energy ginsenosides not studied previously demonstrate a considerable affinity with BACE1, specifically the analogues CHEMBL451292, CHEMBL510371 and CHEMBL503302, with binding energies of -9.6, -8.1 and -7,6 Kcal/mol, respectively. Amino acid residues such as D32, Q73, G230, N233, and R235 were identified as being associated with the formation of hydrogen bonds. Likewise, van der Waals hydrophobic interactions were established by residues L30, G34, S35, T72, F108, I110, I118, T231, T232, R307, K321 and S325, as reported in the study by Choi et al. 2016, which established a strong association between these interactions and the sugars of the molecular structures, which bear a great resemblance to the results obtained for the analogues with higher affinities22. On the other hand, similar interactions with the reference molecules (BACE1 and AChE inhibitors) with common residues identified in the present study and reported previously, such as D32, F108, I118, T232, N233 and K321, were demonstrated, with energy values of -9.8 and -8.3 Kcal/mol for the BACE1 and AChE inhibitors, respectively. It was indicated that the analogues CHEMBL451292 and CHEMBL510371, when interacting with BACE1, demonstrated a similar affinity to the reference drugs, with a higher affinity than donepezil and a similar affinity to Verubecestat, agents that are relevant to the current modern treatments against Alzheimer's disease. Donepezil is used in the usual treatment of the disease in its early and middle phases23 and, despite it having been established that its mechanism of action is based on the inhibition of AchE, our study established that said molecule exhibited binding to active site residues of BACE1, as shown in Figure 2B. On the other hand, Verubecestat is an experimental drug for the treatment of early stage disease, which was developed as an inhibitor of BACE1. During its development it showed interfering activity in the formation of the amyloid cascade and in avoiding the aggregation effect of the β-amyloid plaques. However, during the clinical phase, the study was discontinued due to safety concerns about observed alterations in the liver24,25.\n\nTherefore, the use of computational tools for screening of new therapeutic agents with potential pharmacological activity could contribute to the elucidation at the molecular level of the mechanism of action of ginsenosides, as a group of promising metabolites for the prevention and treatment of Alzheimer's disease.\n\n\nData availability\n\nFigshare: Ginsenoside analogs as possible BACE1 inhibitors. https://doi.org/10.6084/m9.figshare.8408549.v126\n\nThis project contains the following underlying data:\n\n- In silico study ofResults_DOCKED_27LIGANDS.xlsx (analogue and drug structures listed in .xlsx format)\n\nFigshare: Ginsenoside analogs as possible BACE1 inhibitors. https://doi.org/10.6084/m9.figshare.8408549.v126\n\nThis project contains the following extended data:\n\n- 597013_donepezil_out.pdb (donepezil structure in PDB format)\n\n- 51352361_uff_E=743.53_out.pdbqt (verubecestat structure in PDBQT format)\n\n- Structures of the 27 ginsenoside analogues in PDB format\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nUniversitary Corporation Rafael Nuñez and University of Cartagena.\n\n\nReferences\n\nMorley JE, Farr SA, Nguyen AD: Alzheimer Disease. Clin Geriatr Med. 2018; 34(4): 591–601. PubMed Abstract \| Publisher Full Text\n\nXu W, Yu JT, Tan MS, et al.: Cognitive reserve and Alzheimer's disease. Mol Neurobiol. 2014; 51(1): 187–208. PubMed Abstract \| Publisher Full Text\n\nIslam MA, Pillay TS: β-secretase inhibitors for Alzheimer's disease: identification using pharmacoinformatics. J Biomol Struct Dyn. 2019; 37(2): 503–22. PubMed Abstract \| Publisher Full Text\n\nKorolev I: Alzheimer’s Disease: A Clinical and Basic Science Review. Med Student Res J. 2014; 04: 24–33. Reference Source\n\nSingh J, Sabbagh MN, Nair AK, et al.: Treatment of Dementia. Geriatr Neurol. 2014; 76(Suppl V): 556–85.\n\nScarpini E, Scheltens P, Feldman H: Treatment of Alzheimer's disease: current status and new perspectives. Lancet Neurol. 2003; 2(9): 539–47. PubMed Abstract \| Publisher Full Text\n\nLi TSC, Mazza G, Cottrell AC, et al.: Ginsenosides in Roots and Leaves of American Ginseng. J Agric Food Chem. 1996; 44(3): 717–20. Publisher Full Text\n\nKim JH, Yi YS, Kim MY, et al.: Role of ginsenosides, the main active components of Panax ginseng, in inflammatory responses and diseases. J Ginseng Res. 2017; 41(4): 435–43. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKhan S, Tosun A, Kim YS: Ginsenosides as Food Supplements and Their Potential Role in Immunological and Neurodegenerative Disorders. Bioactive Nutraceuticals and Dietary Supplements in Neurological and Brain Disease: Prevention and Therapy. Elsevier Inc.; 2015. 303–309. Publisher Full Text\n\nLeung KW, Wong AS: Pharmacology of ginsenosides: a literature review. Chin Med. 2010; 5: 20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGaulton A, Hersey A, Nowotka M, et al.: The ChEMBL database in 2017. Nucleic Acids Res. 2017; 45(D1): D945–54. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIrwin JJ, Sterling T, Mysinger MM, et al.: ZINC: a free tool to discover chemistry for biology. J Chem Inf Model. 2012; 52(7): 1757–68. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDassault Systemes BIOVIA: Discovery Studio Visualizer 4.5 (2016). 2016.\n\nBerman HM, Westbrook J, Feng Z, et al.: The Protein Data Bank. Nucleic Acids Res. 2000; 28(1): 235–42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPettersen EF, Goddard TD, Huang CC, et al.: UCSF Chimera--a visualization system for exploratory research and analysis. J Comput Chem. 2004; 25(13): 1605–12. PubMed Abstract \| Publisher Full Text\n\nDassault Systèmes BIOVIA: Discovery Studio Visualizer 4.5. 2016.\n\nTrott O, Olson AJ: Software News and Update AutoDock Vina: Improving the Speed and Accuracy of Docking with a New Scoring Function, Efficient Optimization, and Multithreading. J Comput Chem. 2010; 31(2): 455–461. Publisher Full Text\n\nDallakyan S, Olson AJ: Small-molecule library screening by docking with PyRx. Methods Mol Biol. New York 2015: Springer Science+Business Media; 2015; 1263: 243–50. PubMed Abstract \| Publisher Full Text\n\nDeLano WL: PyMOL. Schrödinger; 2019; 1.\n\nBhushan I, Kour M, Kour G, et al.: Alzheimer’s disease: Causes & treatment – A review. Ann Biotechnol. 2018; 1(1): 1002. Reference Source\n\nCoimbra JRM, Marques DFF, Baptista SJ, et al.: Highlights in BACE1 Inhibitors for Alzheimer's Disease Treatment. Front Chem. 2018; 6: 178. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChoi RJ, Roy A, Jung HJ, et al.: BACE1 molecular docking and anti-Alzheimer's disease activities of ginsenosides. J Ethnopharmacol. 2016; 190: 219–30. PubMed Abstract \| Publisher Full Text\n\nMeguro K, Kasai M, Akanuma K, et al.: Donepezil and life expectancy in Alzheimer's disease: a retrospective analysis in the Tajiri Project. BMC Neurol. 2014; 14(1): 83. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEgan MF, Kost J, Voss T, et al.: Randomized Trial of Verubecestat for Prodromal Alzheimer's Disease. N Engl J Med. 2019; 380(15): 1408–20. PubMed Abstract \| Publisher Full Text\n\nKennedy ME, Stamford AW, Chen X, et al.: The BACE1 inhibitor verubecestat (MK-8931) reduces CNS β-amyloid in animal models and in Alzheimer's disease patients. Sci Transl Med. 2016; 8(363): 363ra150. PubMed Abstract \| Publisher Full Text\n\nContreras-Puentes N, Alvíz-Amador A, Mercado-Camargo J: Ginsenoside analogs as possible BACE1 inhibitors. figshare. Dataset. 2019. http://www.doi.org/10.6084/m9.figshare.8408549.v1" }	[ { "id": "52068", "date": "19 Aug 2019", "name": "Zhe Wang", "expertise": [ "Reviewer Expertise Alzheimer's disease", "biochemistry", "cell biology" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nContreras-Puentes et al., searched the database for the analogues of ginsenoside, and identified three compounds that strongly bound to BACE1 by structure simulation. The bindings of the compounds with BACE1 were compared with that of other known BACE1 inhibitors including verubecestat and donepezil, an acethylcholine transferase inhibitor with some inhibitory effect on BACE1 activity. BACE1 residues involved in the bindings and the forces mediating the bindings were characterized.\nThe compounds identified in this paper can be tested for the effects on BACE1 functions, both in vitro and in vivo. While bindings of compounds with enzymes may result in a suppression of enzyme activities, an activating effect cannot be ruled out. If the compounds do inhibit beta-cleavage of APP by BACE1, their applications in AD prevention and therapy could be explored.\nA recent clinical trial showed that verubecestat, a potent BACE1 inhibitor, failed to correct AD symptoms. BACE1 is required for a number of neuronal activities, and the inhibition of BACE1 may cause severe side effects. In such a context, compounds specifically targeting beta-cleavage of APP by BACE1 without affecting other BACE1 functions are desired, and these compounds could be useful for AD prevention. Hence, the compounds identified in this paper could also be tested for specificity.\nIt would be more informative if the authors could analyse the bindings of these compounds with BACE2 that is a close homolog of BACE1, and involved in APP processing and neuronal apoptosis. To our knowledge, all known BACE1 inhibitors also inhibit BACE2, despite of their higher preference for BACE1.\n7.5kcal/mol was set as the energy threshold for compound screening. It would be better to justify this threshold.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "56034", "date": "22 Nov 2019", "name": "Andrew R. Leach", "expertise": [ "Reviewer Expertise drug discovery", "computer modelling and simulation", "cheminformatics", "structure-based design" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nUnfortunately, there are many issues with this paper and the underlying studies, to the extent that the authors really need to rethink the entire programme of research reported.\nThe authors have the very laudable objective, of identifying potential new treatments for Alzheimer's disease. They focus on identifying inhibitors of the BACE1 enzyme. This is a drug target that has been the subject of intense activity over the past 10+ years in Pharma, biotech and academic drug discovery. Many inhibitors of BACE1 have been reported in the literature and in patents together with multiple crystal structures and associated data. A quick search in ChEMBL indicates >8700 inhibitors for human BACE1 and 385 crystal structures in PDBe. Furthermore, multiple compounds have progressed into clinical trials, including several into Phase III. Unfortunately, however, these have been unsuccessful, to the extent that the so-called \"amyloid hypothesis\" is under question.\nOverall, the authors unfortunately do not appear to be aware of, let alone to have made much use, of the available data. They are vague on key aspects of their computational protocols. They do not cite relevant literature, nor do they appear to be aware of key limitations with some of their approaches. There is no experimental validation of their computational predictions. I will now address in more detail some of these issues.\nA quick search using the ChEMBL interface suggests that there are 474 Ginsenoside analogues using the default ChEMBL similarity search method and threshold (85%). This appears to be much more than the authors report (further, \"80-90% similarity\" is too vague a description of their approach).\n\nThese 474 analogues have >4000 reported activities in ChEMBL. However, none of these are against BACE1. Thus, there is no evidence from the literature/ChEMBL that ginsenoside or its analogues bind to BACE1 (nor would there be any particular reason to believe that they are active at this target, on the basis of their chemical structure)\n\nTwo other drugs were selected, one a known BACE1 inhibitor and the other an ACHe inhibitor. I can understand the logic for including the former, but why was the ACHe inhibitor included? I cannot find any data in ChEMBL to suggest that donepezil has any activity against BACE1. Is this therefore included as a \"known inactive\" to provide a comparison?\n\nThere is a crystal structure available of verubecestat bound to BACE1 (PDB code 5hu1). The authors seem unaware of this fact and choose to use a very early BACE1 structure for their docking studies. They make no reference to this known structure when presenting the results of their own docking studies for this molecule. Moreover, the docking mode they obtain is completely different to the known crystal structure... key interactions with the catalytic asp residues are not present and indeed the whole binding mode looks to be wrong. If the authors cannot demonstrate that their docking protocol reproduces known structures I would have serious concerns about all of the subsequent work and the conclusions drawn.\n\nA long-standing, and still largely unsolved problem in \"docking and scoring\" is that the scoring methods in current use are very rarely able to reproduce experimentally measured binding affinities. I would thus have absolutely no confidence in the \"binding energies\" reported by the authors. The authors provide no supporting evidence nor citations to previous work that validates their docking and scoring protocol in the context of known experimental data. The calculated free energy differences between the known inhibitor verubecestat and the ACHe inhibitor donepezil from their docking calculations is ~1.5kcal/mol. This would translate to just over a 10-fold difference in binding constant. Given that verubecestat has reported pIC50 affinities in the 8.3-9.4 range, this would suggest that donepezil should be a reasonable inhibitor of BACE1. There is no literature evidence that I can find to support this. Given all of this, I can only conclude that all of the reported \"free energies of binding\" in this paper are wholly erroneous and moreover the docking modes on which they are based are similarly suspect.\n\nIrrespective of the issues described above, any predictions made using computational methods should be tested in the laboratory. Thus, we should not be seeing papers such as this which only contain computational predictions (and certainly not in a reputable journal such as f1000). Rather, we need to see experimental validation, by testing some or all of the molecules described in an appropriate assay (possibly also supported by crystal structure determination).\n\nThere are also a number of smaller problems with the paper including typographical issues (e.g. the caption to figure 2 defers to \"BASE1\"). In view of the above significant and major problems I will not capture any more of these in this report.\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [] } ]	1	https://f1000research.com/articles/8-1169
https://f1000research.com/articles/8-1155/v1	22 Jul 19	{ "type": "Study Protocol", "title": "Evaluation of benefits and harms of surgical treatments for post-radical prostatectomy urinary incontinence: a systematic review and meta-analysis protocol", "authors": [ "Roselyne Choiniere", "Patrick O. Richard", "Melanie Morin", "Le-Mai Tu", "Gordon H. Guyatt", "Philippe D. Violette", "Roselyne Choiniere", "Patrick O. Richard", "Melanie Morin", "Le-Mai Tu", "Gordon H. Guyatt" ], "abstract": "Background: Post-radical prostatectomy urinary incontinence (PPI) is a frequent and feared complication that can affect approximately 25% of patients. Between 1 and 10% of patients suffering from PPI will require surgery. The effectiveness of the available surgical interventions has only been compared in a few randomized controlled trials and the available reviews have important limitations regarding both benefits and harms that make them insufficient to inform decision-making. The aim of the study is to provide systematic summaries of benefits and harms of contemporary surgical treatment options for PPI through systematic review and meta-analysis using GRADE methodology and reporting in accord with the PRISMA-P statement. Methods: Studies pertaining to bulking agents, male synthetic slings, compressive balloon systems (ProACT) or artificial urinary sphincters (AUS) used for the treatment of patients suffering from PPI will be included. A systematic search will be conducted using the OVID and PubMED platforms in MEDLINE, Embase and Cochrane databases, and reference lists of relevant reviews and guidelines. Trained independent reviewers will conduct study selection and data extraction. Outcomes will include the number of pads used per day, the 24-h pad weight test, the Patient Global Impression of Improvement (PGI-I) and the Incontinence Quality of Life (IQOL) as possible benefits and the reoperations, the Clavien-Dindo complications and the other reported adverse events as the harms. When possible, pooled analyses will be completed. Risk of bias will be assessed using the CLARITY tools and a new tool for the before-and-after studies without a control group. Finally, study heterogeneity will be assessed, publication bias will be evaluated with funnel plots and quality of evidence rated for each outcome. Discussion: Our study will address patient-important outcomes and will be useful in clinical decision-making as well as identifying key elements for future research. Study registration: PROSPERO: CRD42018073923 05/12/2018", "keywords": [ "meta-analysis", "systematic review", "urinary incontinence", "prostatectomy", "protocol" ], "content": "Introduction\n\nRadical prostatectomy is one of the mainstays of treatment for prostate cancer. However, radical prostatectomy is associated with significant morbidity and complications1–3. A common and feared complication is persistent post-radical prostatectomy urinary incontinence (PPI)4. Following radical prostatectomy, patients suffering from PPI usually report a gradual continence improvement with conservative measures within 12 months postoperatively5,6. However, reports show that, 1-year after a radical prostatectomy, up to 25% of patients will suffer from some degree of PPI7,8. This can have significant impact on quality of life and may influence social relationships, emotional health and physical exercise9,10.\n\nConservative treatments including lifestyle modifications, bladder training and pelvic floor physiotherapy are recommended as the first-line therapy, but a significant proportion of patients with PPI will seek surgical treatment in the long term due to persistent PPI11. The historic gold standard operation for PPI is the placement of an artificial urinary sphincter (AUS)12. Indeed, a multicenter population-based Canadian study on 25 346 men showed that 2.6%, 3.8% and 4.8% of patients received AUS or a sling at 5, 10 and 15 years following a radical prostatectomy respectively13. Similarly, a nationwide study using the American College of Surgeons National Cancer Database has shown that the incidence of AUS post-radical prostatectomy varied from 1 to 10% (mean 6%) totalling the use of 4 426 AUS for 79 900 Radical prostatectomies (RPs) annually in the United States14.\n\nIn contemporary practice, alternative surgical treatments for PPI include four main interventions: bulking agents, male synthetic slings, compressive balloon systems (also known as ProACTTM) and AUS, the current gold standard.\n\nChoosing the adequate device is a challenge for patients and physicians often due to limited understanding of relative efficacy and harms for each intervention. A Cochrane review attempted to summarize the subject in 2014, but was limited due to the paucity of randomized controlled trials (RCTs), in fact, citing one small trial with only 45 patients. Thus, it provided few meaningful conclusions and minimal clinical guidance12,15.\n\nIn 2017, Chen et al. published a systematic review and meta-analysis addressing the subject16. By necessity, these authors included data from observational studies in their assessment and did provide an assessment of efficacy for some procedures for PPI. However, this work remains incomplete in several important aspects and has a limited use in clinical decision-making. First, the authors restricted their search to male synthetic slings and AUS. They did not include the bulking agents and they did not report the compressive balloon systems as a separate category. This limits the information on potential treatment choices patients may consider. Second, the authors limited their inclusion criteria to RCTs and prospective observational studies. Unfortunately, the majority of studies published on this topic have a retrospective design. Consequently, the authors were unable to adequately assess harms due to lack of data. Third, although it is believed that certain patient’s characteristics may impact the outcomes of PPI surgeries such as baseline severity of incontinence and history of pelvic radiotherapy, the authors did not ascertain nor discuss the influence of these characteristics on the treatment efficacy17,18. Fourth, the authors did not specify the time frames used in the analyses of efficacy which lessens the clinical interpretability of their findings given the expected variation of efficacy, harms and quality of life outcomes over time19,20. Lastly, the majority of data included was nearly a decade old and may not represent contemporary practice.\n\nOverall, this important knowledge gap limits our ability to engage with our patients and make informed decisions about treatment. Many questions remain unanswered regarding the pertinent trade-offs for each device, specifically the potential impact on continence, quality of life and adverse events. We propose to fill this gap by conducting a systematic review and, where possible, a meta-analysis to provide summaries of patient-important outcomes to inform clinical decision-making.\n\n\nMethods\n\nThis protocol adheres to the recommendations of the Preferred Reporting Items for Systematic Reviews and Meta-analysis Protocols (PRISMA-P) statement, the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach and the Cochrane Handbook methodology for Systematic Reviews of Interventions (Version 5.1.0)21–23. This protocol was registered in PROSPERO CRD42018073923 on December 5th, 2018.\n\nThis research was done without patient involvement.\n\n\nEligibility criteria\n\nRCTs and observational studies that initially enrolled a minimum of 50 patients suffering from PPI per group will be included in the review. Studies without a comparison group will also be included. Primary studies published after January 1st, 1997 will be included. If more than one study publishes results from the same cohort, the most recent results will be included. There will be no restrictions based on language or country of origin.\n\nStudies that involved adult men suffering from PPI after radical prostatectomy will be included in the review. There will be no restrictions based on radical prostatectomy surgical technique, prostate cancer stages, baseline severity of PPI and history of prior pelvic radiotherapy or previous failed corrective incontinence surgery.\n\nAny studies reporting on a surgical intervention meant to cure PPI using either an implantation of a device (i.e. male synthetic slings, compressive balloon systems and AUS) or a bulking agent with or without a comparison group will be considered for inclusion. Studies in which the continence surgery was completed simultaneously with another concomitant intervention (i.e. radical prostatectomy, penile prosthesis or any other intervention) will be excluded.\n\n\nOutcome measures\n\nStudies reporting at least one of the following outcomes of benefits or harms will be included. Data obtained with more than 20% lost at follow-up will be excluded from the analysis.\n\nBenefits\n\nPrimary outcome\n\n1. Cure, improvement and failure rates defined by the number of pads per day.\n\na. Several definitions of treatment success with the number of pads per day are used by authors. We will analyze results according to the following definitions: patients will be considered as 1) cured, if they wear no pad per day to a maximum of a safety pad or 1 pad per day; 2) improved, if they report a reduction of ≥50% of the number of pads per day and/or wear 2 or fewer pads per day; and 3) treatment failures, if they are not cured or improved and/or are wearing 3 or more pads per day.\n\nSecondary outcomes\n\nEfficacy\n\n2. Cure, improvement and failure rates at the 24-h pad weight test.\n\na. The 24-h pad weight test also has different definitions of treatment success. As such, cure will be defined as 24-h pad weight test of 0 g to <10 g per day. Improvement will be characterized as reduction of ≥50% of urine loss per day. Patients that are neither cured nor improved will be considered as treatment failures.\n\n3. The number of pads per day at follow-up.\n\n4. Mean reduction of weight of urine loss at the 24-h pad weight test from baseline.\n\n5. Mean impression of improvement of incontinence will be evaluated by the Patient Global Impression of Improvement (PGI-I)24.\n\nQuality of life\n\n6. Improvement of quality of life reported by validated questionnaires such as the Incontinence-Quality of Life (IQOL), which addresses the patient’s feelings about his condition and quantifies bother on daily activities25.\n\nHarms and adverse events\n\nPrimary outcome\n\n7. Reoperation rates.\n\na. Reoperations include all direct causes of a secondary surgical procedure such as surgical revisions, explantation and implantation of a subsequent device.\n\nSecondary outcomes\n\n8. Short term perioperative complications as reported according to the Clavien-Dindo classification26.\n\na. The grades I and II will be considered as minor complications and grades III, IV and V will be considered as major complications.\n\n9. Long-term adverse events as defined as the rates of revision (secondary operation) and explantation (removal) of the surgical device27,28.\n\n10. Additional reported adverse events will also be documented.\n\n\nInformation sources\n\nAn extensive and systematic electronic search will be performed for the following databases: MEDLINE via PubMED and The Cochrane Library for the relevant publications published from January 1st, 1997. In addition, reference lists of relevant articles such as review articles and guidelines will be manually screened for other eligible studies.\n\n\nSearch strategy\n\nThis research will be achieved using specific keywords combinations and Medical Subject Headings (MeSH) terms previously defined by our content expert (LMT) and an experienced research librarian for the different platforms. An example of search strategy string is available as Extended data29. The identified publications will be managed with the Zotero 5.0 software and duplicates will be removed.\n\n\nStudy records\n\nThroughout the systematic review, data will be managed with the software Zotero 5.0 and Excel sheets.\n\n1. Title and abstract screening: A first screening will be performed by reviewing the title and abstract of each of the identified publication. All reviewers will undergo training before commencing the screening phase. Training will consist of titles and abstracts screening of 15 articles. Trainees will be considered to have successfully completed training if they obtain a concordance of 90% or higher on the standardized test prepared by the first author (RC). Training will be redone if necessary. This first screening will be completed by independent teams of two reviewers. Disagreements between team members will be resolved by discussion among themselves or by a third reviewer (RC), if required. The remaining selected studies will be used for the next step.\n\n2. Full text screening: A second screening will also be performed by independent reviewers to complete study selection. As in the first screening phase, reviewers will be previously trained for the specific criteria of this second step with a screening test of 15 articles. Trainees will be considered to have successfully completed training if they obtain a concordance of 90% or higher. Training will be redone if necessary. Reviewers will perform a full-read selection and record the main reason for excluding each study in line with the PRISMA Collaboration21. Disagreements will be resolved among reviewers’ teams or by a third reviewer (RC), if required. The selection process will be presented in a PRISMA flow diagram.\n\nFollowing study selection, data extraction will be performed on the eligible articles in duplicate. A specific data collection form was be created with Excel software and pilot tested (see Extended data29). Trained independent reviewers will execute a thorough read of their assigned articles and complete the form. All extracted data will be reviewed and compared between both reviewers of each team. Disagreements of extracted data between reviewers will be resolved among themselves and, if needed, by a third reviewer (PR), a clinician methodologist.\n\nThe data collection form is available as Extended data29 and consists of 4 main sections:\n\n1. General information,\n\nFirst author’s last name, year of publication, country of origin of the first author, study design, number of surgeons and centers, enrollment and intervention period, initial number of patients per intervention group, number of patients with PPI, loss of patients at follow-up, length of follow-up, period between radical prostatectomy and continence intervention,\n\n2. Patient characteristics,\n\nAge, body mass index, baseline severity of urinary incontinence, history of pelvic irradiation,\n\n3. Intervention and comparison,\n\nType of device, name of the device, comparative device,\n\n4. Outcome results for efficacy, quality of life and adverse events before and after the intervention.\n\nNumber of pads per day, 24-h pad weight test, the improvement questionnaire, the quality of life questionnaire and adverse events.\n\nRisk of bias will be assessed for each individual study using instruments appropriate to study design as described below. Pairs of trained independent reviewers will perform risk of bias assessment. Disagreements will be solved among reviewers and, if needed, by a third reviewer (PR or PV). The risk of bias assessment will be represented in coloured graphs.\n\nStudies will be defined as having a lower risk of bias if they have ≤2 \"probably high risk of bias\" domains and no \"definitely high risk of bias\" domain. Studies with 3 or more \"probably high risk of bias\" domains and/or one or more domains associated with a \"definitely high risk of bias\" will be considered as having a higher risk of bias. In analysis of pooled estimates, studies with a lower risk of bias will be preferred to represent the overall effect if a statistical difference is found between the subgroup pooled estimates.\n\nFor RCTs, we will use a modified version of one of the Cochrane Risk of Bias instruments, which ranks risk of biases as definitely high to definitely low risk of bias. This tool addresses the biases associated with allocation sequencing, concealment of allocation, blinding, missing data and selective reporting30.\n\nRisk of bias of cohort studies will be assessed with the CLARITY risk of bias tools. Patient selection, exposure and outcome assessment, as well as, missing data will be evaluated30.\n\nWe anticipate that much of the evidence may come from before-and-after studies without a control group. This design poses specific challenges in assessing risk of bias. We performed a systematic search of the literature in PubMed, Google Scholar, websites of prominent methodological organizations in evidence-based medicine and relevant systematic reviews to identify appropriate risk of bias instruments for before-and-after studies31,32. We were not able to identify an instrument that we judged fully satisfactory. Many instruments combined quality and reporting issues with assessment of risk of bias and failed to identify issues relevant to this particular review.\n\nTherefore, we developed a before-and-after risk of bias instrument in two stages. Initially, we applied the domains from well-established risk of bias instruments such as those developed by the CLARITY research group, the ROBINS-I, the National Institute of Health before and after tool and an instrument developed by the Johanna Briggs Institute30,33–35. Then, we reduced the number of questions to 4 domains that we believe capture the essential risk of biases items for before-and-after studies36. This risk of bias instrument is available as Extended data29.\n\nDescriptive data will be used to characterize our study population. Characteristics of interest will consist of age, body mass index and timing between radical prostatectomy and incontinence surgery.\n\nQuantitative analyses for benefits and harms outcomes will be performed according to study design. For RCTs and for cohort studies of comparable populations, we will use a random effects model to calculate the pooled estimates of effect size. These pooled estimates will then be represented in forest plots. For before-and-after studies with no control group, each intervention will be analyzed separately to obtain pooled estimates which will be compared to the AUS in a narrative summary. The studies will be weighted using the inverse variance that incorporates the variance and the number of patients. This method will be used to compensate for varying sample size among the studies to allow larger studies to have more weight in the analysis37. Mean differences for continuous outcomes and proportions for dichotomous outcomes with 95% confidence interval (CI) will be pooled. Continuous outcomes will be the change in the 24-hour pad weight test and the improvement and quality of life questionnaires. Dichotomous outcomes for benefits and harms will be the cure, improved and failed rates of treatment, reoperations rates and Clavien-Dindo complications.\n\nStudies will be pooled by types of intervention: bulking agents, male synthetic slings, compressive balloon systems and AUS. We will provide summary data for each intervention type.\n\nPooled analyses will be performed according to length of follow-up. For all outcomes, with the exception of the Clavien-Dindo complications, a main analysis will be conducted using data acquired from 6 months to less than 36 months, preferring data from the follow-up closest to 12 months. For our primary outcomes of benefits and harms, a secondary analysis will be performed using the longest data available that has been acquired at 36 months of follow-up or more.\n\nFinally, summary statistics will be provided and a narrative report of the findings will be completed. The meta-analyses will be performed in collaboration with an experienced specialized statistician.\n\nFor results reported without a variance measure, we will use the Wan et al. method to calculate missing data38. For continuous outcomes, when the standard deviation (SD) of change from before to after the intervention is not reported, we will use an imputation method from the Cochrane Handbook assuming correlations of 0.1, 0.5 and 0.9 to calculate the SD of change from SD at baseline and SD at assessment followed by sensitivity analyses23.\n\nIf possible, data reported in graphs will be extracted when not reported otherwise. In addition, when not specified, we will assume the number of participants at assessment by using the initial number of participants.\n\nHeterogeneity between studies will be measured using visual inspection of the forest plots and χ2 test.\n\nThe I2 statistic will be reported to show the percentage of variability that is due to true differences between studies (heterogeneity) rather than a sampling error (chance). In agreement with the Cochrane and GRADE handbooks, 0 to 40% will be interpreted as ‘’might not be important“, 30 to 60% as ‘’may represent moderate heterogeneity”, 50 to 90% as ‘’may represent substantial heterogeneity and 75 to 100% as ‘’considerable heterogeneity”23.\n\nThe best available data will be used to provide accurate evidence summary. Within the GRADE approach, outcomes will be reported regardless of I2 results and plausible sources of heterogeneity will be explored with transparency as described below39.\n\nHeterogeneity may also be explained by some aspects from the studies such as risk of bias, length at follow-up and patients’ preoperative characteristics.\n\nWe will perform subgroup analysis by comparing studies rated as having a higher risk of bias to those having lower risk of bias as a means of investigating heterogeneous results.\n\nWe believe another explanation for heterogeneity could be the length of follow-up. Our hypothesis is that efficacy and quality of life outcomes will have better results in the shorter term than in the longer term19,20. We will perform a subgroup analysis for efficacy and quality of life outcomes comparing results obtained at a shorter term (6 to 12 months) and a longer term (more than 12 months).\n\nSeveral factors need to be considered before choosing a device. Some patient characteristics may influence efficacy outcomes, such as the cure, improvement and failure rates at the number of pads per day. First, we hypothesize that patients who received pelvic radiotherapy will have worse outcomes than non-irradiated patients due to fibrosis and scarring40–42. Second, we hypothesize that patients suffering from severe PPI will have worse outcomes following continence surgery than patients suffering from mild and moderate PPI43,44. If there are data available, subgroup analyses will be performed.\n\nWe will detect reporting bias using funnel plots if more than 10 studies are included in any meta-analysis. An asymmetry test will be performed using the Egger method45,46.\n\nQuality of the body of evidence for each outcome will be assessed by using the GRADE approach22. This approach considers the following factors: study design, limitations in study design and implementation, indirectness of evidence, unexplained heterogeneity or inconsistency of results, imprecision of results and publication bias. The latter factor will be assessed by a funnel plot where applicable. Criteria for applicability are described in Publication bias section above.\n\nThe results of the study will be submitted for publication to a peer-reviewed journal. We also intend to share the results at relevant national and international conferences.\n\nThis project is ongoing. Study selection is now completed. Data extraction, risk of bias assessment and data analysis have begun, but are not completed.\n\n\nDiscussion\n\nOur study is intended to be rigorous, systematic and transparent in agreement with the PRISMA-P statement, the GRADE approach and the Cochrane Handbook. Our criteria are precise and explicit and our searches are comprehensive. Our team is composed of content experts, methodologists, a specialized and experienced librarian and a biostatistician specialized in meta-analyses. Moreover, study selection, data extraction and risk of bias assessment will be performed by trained independent teams of reviewers. Data and disagreements will be reviewed by a third reviewer to address discrepancies.\n\nNevertheless, we anticipate that our study will have some limitations. First, we anticipate the paucity of RCT data. There is a possibility most data will come from observational studies that may have lower reporting standards and affect the quality of the available evidence. Second, many of the observational studies will probably use a before-and-after study design that presents specific biases. To specifically evaluate the quality of these studies, we had to create a new risk of bias assessment instrument, given the lack of an adequate one in the literature. Lastly, our study results will be based on published data and may not reflect unreported or rarely reported harms. Although, publication bias will be assessed and discussed, the possibility of negative studies not being published cannot be excluded.\n\nNevertheless, in spite of these limitations, our study addresses patient-important outcomes and will be useful in decision-making. Physicians will be able to consult our review and obtain the best current estimates. Our data will be used in guidelines and recommendations to provide guidance to better practices. Also, this meta-analysis will identify the key points for future research on the subject. Finally, this study will also help prepare and guide a future large-scale RCT.\n\nThis study protocol is registered in PROSPERO CRD42018073923. Available from: http://www.crd.york.ac.uk/PROSPERO/display_record.php?ID=CRD42018073923\n\n\nData availability\n\nNo data is associated with this article.\n\nOpen Science Framework: Evaluation of benefits and harms of surgical treatments for post-radical prostatectomy urinary incontinence: a systematic review and meta-analysis protocol. https://doi.org/10.17605/OSF.IO/PVNWX29\n\nThis project contains the following extended data:\n\nSearch Strategy String_RChoiniere.docx (A search strategy string used for one of the databases for the systematic review)\n\nData Collection Sheet_RChoiniere.xlsx (The data collection sheet used for data extraction in the study)\n\nRisk of bias Instrument - Before After studies_RChoiniere.docx (The risk of bias instrument created by the Male Incontinence Research Group for before-after studies)\n\nOpen Science Framework: PRISMA-P checklist for \"Evaluation of benefits and harms of surgical treatments for post-radical prostatectomy urinary incontinence: a systematic review and meta-analysis protocol\". https://doi.org/10.17605/OSF.IO/PVNWX29\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nÁvila M, Patel L, López S, et al.: Patient-reported outcomes after treatment for clinically localized prostate cancer: A systematic review and meta-analysis. Cancer Treat Rev. 2018; 66: 23–44. PubMed Abstract \| Publisher Full Text\n\nSeo HJ, Lee NR, Son SK, et al.: Comparison of Robot-Assisted Radical Prostatectomy and Open Radical Prostatectomy Outcomes: A Systematic Review and Meta-Analysis. Yonsei Med J. 2016; 57(5): 1165–77. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTang K, Jiang K, Chen H, et al.: Robotic vs. Retropubic radical prostatectomy in prostate cancer: A systematic review and an meta-analysis update. Oncotarget. 2017; 8(19): 32237–57. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHoyland K, Vasdev N, Abrof A, et al.: Post-radical prostatectomy incontinence: etiology and prevention. Rev Urol. 2014; 16(4): 181–8. PubMed Abstract \| Free Full Text\n\nFicarra V, Novara G, Artibani W, et al.: Retropubic, laparoscopic, and robot-assisted radical prostatectomy: a systematic review and cumulative analysis of comparative studies. Eur Urol. 2009; 55(5): 1037–63. PubMed Abstract \| Publisher Full Text\n\nLepor H, Kaci L, Xue X: Continence following radical retropubic prostatectomy using self-reporting instruments. J Urol. 2004; 171(3): 1212–5. PubMed Abstract \| Publisher Full Text\n\nDonovan JL, Hamdy FC, Lane JA, et al.: Patient-Reported Outcomes after Monitoring, Surgery, or Radiotherapy for Prostate Cancer. N Engl J Med. 2016; 375(15): 1425–37. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nResnick MJ, Koyama T, Fan KH, et al.: Long-term functional outcomes after treatment for localized prostate cancer. N Engl J Med. 2013; 368(5): 436–45. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSandhu JS: Treatment options for male stress urinary incontinence. Nat Rev Urol. 2010; 7(4): 222–8. PubMed Abstract \| Publisher Full Text\n\nHamdy FC, Donovan JL, Lane JA, et al.: 10-Year Outcomes after Monitoring, Surgery, or Radiotherapy for Localized Prostate Cancer. N Engl J Med. 2016; 375(15): 1415–24. PubMed Abstract \| Publisher Full Text\n\nAnderson CA, Omar MI, Campbell SE, et al.: Conservative management for postprostatectomy urinary incontinence. Cochrane Database Syst Rev. 2015; 1: CD001843. PubMed Abstract \| Publisher Full Text\n\nSilva LA, Andriolo RB, Atallah AN, et al.: Surgery for stress urinary incontinence due to presumed sphincter deficiency after prostate surgery. Cochrane Database Syst Rev. 2014; (9): CD008306. PubMed Abstract \| Publisher Full Text\n\nNam RK, Herschorn S, Loblaw DA, et al.: Population based study of long-term rates of surgery for urinary incontinence after radical prostatectomy for prostate cancer. J Urol. 2012; 188(2): 502–6. PubMed Abstract \| Publisher Full Text\n\nReynolds WS, Patel R, Msezane L, et al.: Current use of artificial urinary sphincters in the United States. J Urol. 2007; 178(2): 578–83. PubMed Abstract \| Publisher Full Text\n\nImamoglu MA, Tuygun C, Bakirtas H, et al.: The comparison of artificial urinary sphincter implantation and endourethral macroplastique injection for the treatment of postprostatectomy incontinence. Eur Urol. 2005; 47(2): 209–13. PubMed Abstract \| Publisher Full Text\n\nChen YC, Lin PH, Jou YY, et al.: Surgical treatment for urinary incontinence after prostatectomy: A meta-analysis and systematic review. PLoS One. 2017; 12(5): e0130867. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCrivellaro S, Morlacco A, Bodo G, et al.: Systematic review of surgical treatment of post radical prostatectomy stress urinary incontinence. Neurourol Urodyn. 2016; 35(8): 875–81. PubMed Abstract \| Publisher Full Text\n\nRadadia KD, Farber NJ, Shinder B, et al.: Management of Postradical Prostatectomy Urinary Incontinence: A Review. [Review]. Urology. 2018; 113: 13–9. PubMed Abstract \| Publisher Full Text\n\nLi H, Gill BC, Nowacki AS, et al.: Therapeutic durability of the male transobturator sling: midterm patient reported outcomes. J Urol. 2012; 187(4): 1331–5. PubMed Abstract \| Publisher Full Text\n\nZuckerman JM, Edwards B, Henderson K, et al.: Extended outcomes in the treatment of male stress urinary incontinence with a transobturator sling. Urology. 2014; 83(4): 939–45. PubMed Abstract \| Publisher Full Text\n\nMoher D, Shamseer L, Clarke M, et al.: Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev. 2015; 4: 1. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHultcrantz M, Rind D, Akl EA, et al.: The GRADE Working Group clarifies the construct of certainty of evidence. J Clin Epidemiol. 2017; 87: 4–13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThe Cochrane Collaboration: Cochrane Handbook for Systematic Reviews of Interventions. Version 5.1.0. Higgins J Green S editors. 2011. Reference Source\n\nViktrup L, Hayes RP, Wang P, et al.: Construct validation of patient global impression of severity (PGI-S) and improvement (PGI-I) questionnaires in the treatment of men with lower urinary tract symptoms secondary to benign prostatic hyperplasia. BMC Urol. 2012; 12: 30. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWagner TH, Patrick DL, Bavendam TG, et al.: Quality of life of persons with urinary incontinence: development of a new measure. Urology. 1996; 47(1): 67–71; discussion 71–2. PubMed Abstract \| Publisher Full Text\n\nClavien PA, Barkun J, de Oliveira ML, et al.: The Clavien-Dindo classification of surgical complications: five-year experience. Ann Surg. 2009; 250(2): 187–96. PubMed Abstract \| Publisher Full Text\n\nAbrams P, Cardozo L, Wagg A, et al.: Incontinence. 6th International Consultation on Incontinence. 6th ed. Tokyo: International Continence Society; (6th International Consultation on Incontinence). 2016; 2636. Reference Source\n\nCordon BH, Singla N, Singla AK: Artificial urinary sphincters for male stress urinary incontinence: current perspectives. Med Devices (Auckl). 2016; 9: 175–83. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChoiniere R: Extended data: Evaluation of Benefits and Harms of Surgical Treatments for Post Radical Prostatectomy Urinary Incontinence: A Systematic Review and Meta-analysis Protocol, Roselyne Choiniere. 2019. http://www.doi.org/10.17605/OSF.IO/PVNWX\n\nEvidence Partners, CLARITY Group: Methodological Resources. Systematic Review and Literature Review Software by Evidence Partners. [cited 2018 Jan 26]. Reference Source\n\nSanderson S, Tatt ID, Higgins JP: Tools for assessing quality and susceptibility to bias in observational studies in epidemiology: a systematic review and annotated bibliography. Int J Epidemiol. 2007; 36(3): 666–76. PubMed Abstract \| Publisher Full Text\n\nZeng X, Zhang Y, Kwong JS, et al.: The methodological quality assessment tools for preclinical and clinical studies, systematic review and meta-analysis, and clinical practice guideline: a systematic review. J Evid Based Med. 2015; 8(1): 2–10. PubMed Abstract \| Publisher Full Text\n\nSterne JA, Hernán MA, Reeves BC, et al.: ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions. BMJ. 2016; 355: i4919. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nU.S. Department of Health & Human Services: Study Quality Assessment Tools \| National Heart, Lung, and Blood Institute (NHLBI). [cited 2018 Jan 26]. Reference Source\n\nJBI Database of Systematic Reviews and Implementation Reports: Critical Appraisal Tools - JBI. 2015. Reference Source\n\nGuyatt GH, Oxman AD, Vist G, et al.: GRADE guidelines: 4. Rating the quality of evidence--study limitations (risk of bias). J Clin Epidemiol. 2011; 64(4): 407–15. PubMed Abstract \| Publisher Full Text\n\nHaidich AB: Meta-analysis in medical research. Hippokratia. 2010; 14(Suppl 1): 29–37. PubMed Abstract \| Free Full Text\n\nWan X, Wang W, Liu J, et al.: Estimating the sample mean and standard deviation from the sample size, median, range and/or interquartile range. BMC Med Res Methodol. 2014; 14(1): 135. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGuyatt GH, Oxman AD, Kunz R, et al.: GRADE guidelines: 7. Rating the quality of evidence--inconsistency. J Clin Epidemiol. 2011; 64(12): 1294–302. PubMed Abstract \| Publisher Full Text\n\nGregori A, Romanò AL, Scieri F, et al.: Transrectal ultrasound-guided implantation of Adjustable Continence Therapy (ProACT): surgical technique and clinical results after a mean follow-up of 2 years. Eur Urol. 2010; 57(3): 430–6. PubMed Abstract \| Publisher Full Text\n\nWestney OL, Bevan-Thomas R, Palmer JL, et al.: Transurethral collagen injections for male intrinsic sphincter deficiency: the University of Texas-Houston experience. J Urol. 2005; 174(3): 994–7. PubMed Abstract \| Publisher Full Text\n\nGuillaumier S, Solomon E, Jenks J, et al.: Radiotherapy is associated with reduced continence outcomes following implantation of the artificial urinary sphincter in men with post-radical prostatectomy incontinence. Urol Ann. 2017; 9(3): 253–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeruth J, Waltregny D, de Leval J: The inside-out transobturator male sling for the surgical treatment of stress urinary incontinence after radical prostatectomy: midterm results of a single-center prospective study. Eur Urol. 2012; 61(3): 608–15. PubMed Abstract \| Publisher Full Text\n\nRouprêt M, Misraï V, Gosseine PN, et al.: Management of stress urinary incontinence following prostate surgery with minimally invasive adjustable continence balloon implants: functional results from a single center prospective study. J Urol. 2011; 186(1): 198–203. PubMed Abstract \| Publisher Full Text\n\nEgger M, Davey Smith G, Schneider M, et al.: Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997; 315(7109): 629–34. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSterne JA, Egger M: Funnel plots for detecting bias in meta-analysis: guidelines on choice of axis. J Clin Epidemiol. 2001; 54(10): 1046–55. PubMed Abstract \| Publisher Full Text" }	[ { "id": "52771", "date": "16 Sep 2019", "name": "Dean S Elterman", "expertise": [ "Reviewer Expertise Suregon-Investigator in Urology with Graduate degree in Clinical epidemiology and health services research." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe proposed protocol for a systematic review and meta-analysis on surgical treatments for post-radical prostatectomy incontinence is an important endeavour given the availability of surgical options beyond the artificial urinary sphincter. The authors appreciate the paucity of prospective studies to date, let alone randomized controlled trials on this topic. Thus, retrospective and observational studies are included in this review.\n\nThe authors have a practical approach in defining cure (primary outcome) as “patients wearing no pads per day to a maximum of a safety pad or 1 pad per day\". This definition is more clinically relevant and applicable to daily practice. Their proposed secondary outcomes capture the various endpoints typically used in studies related to incontinence, such as 24-hr pad weights and validated questionnaires. Other data items proposed for collection also encompasses important factors of interest such as baseline severity, BMI, age and history of pelvic radiation to name a few.\n\nWith regards to the proposed data analysis component, the authors have anticipated the need for assessing studies with no control groups. They have outlined the risk of bias instrument that they will be using to assess this issue which was modified from validated instruments and reduced to 4 domains of interest (non-validated). However, I wonder if question #3 of the questionnaire “were outcome assessors blinded and independent to the intervention?” would be able to discern the risk of bias with the study as I anticipate a large proportion of which would the assessors would not be blinded. Choosing aspects of validated methods to ascertain bias does not guarantee accurate assessment of bias. I don't know if taking a portion of a bias tool and applying it invalidates the utility of the tool.\n\nI look forward to seeing the results of this undertaking given the potential impact it may have on patient care.\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Not applicable", "responses": [] }, { "id": "75395", "date": "25 Nov 2020", "name": "Bilal Chughtai", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors of this article have put forth a well thought out and sound methodology for a systematic review and meta analysis of a very important topic in urology. The authors have laid out clear criteria for studies to be included in this and their rationale for including studies beyond randomized control trials. Given the limited data on surgical treatment options for urinary incontinence, we agree with their study design.\nThe authors have outlined a comprehensive list of primary and secondary outcomes, which are appropriate for this type of study. While quantifying the degree of incontinence can be challenging, the authors have proposed good objective measures including number of pads used per day and 24 hour pad weight. Additionally, subjective improvement will be measured with the PGI-I and IQOL, with the IQOL measuring the patient’s perspective of bother, a viewpoint which is critical. Additionally, the authors proposed understanding reoperation rates and adverse events, topics of significant interest to both patients and urologists. Our only recommendation to the authors is to include a functional status of patients undergoing the procedures. It is well understood that to utilize an artificial urinary sphincter, the patients need appropriate dexterity. Therefore, it would be helpful to include this (perhaps as an EGOG status) when capturing patient demographics.\nOne of the keys to a successful systematic review and meta analysis is the ability to assess for biases. Well known tools will be utilized by the authors including the Cochrane risk of bias tool and the CLARITY tool. However, for before and after studies, the authors have proposed a novel method of assessing bias by adapting other tools into concise questions. While we applaud the innovative efforts of the authors, there is some concern about utilizing a non-validated modality for assessing bias, particularly for before and after study designs. While we understand the authors’ motivation, we suggest considering a way to validate this new tool prior to implementation to enhance the quality of the systematic review and meta analysis.\nOverall, this is a very strong study design and adheres to many of the standardized recommendations of a systematic review and meta-analysis. We applaud the authors for studying a very important and interesting topic within urologic prosthetics. We look forward to seeing the results of the study.\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1155
https://f1000research.com/articles/8-1153/v1	22 Jul 19	{ "type": "Study Protocol", "title": "Protocol for a factorial randomised controlled trial, embedded within WHiTE 8 COPAL, of an Enhanced Trainee Principal Investigator Package and Additional Digital Nudge to increase recruitment rates", "authors": [ "Nickil Agni", "Caroline Fairhurst", "Catriona McDaid", "Mike Reed", "David J. Torgerson", "Caroline Fairhurst", "Catriona McDaid", "Mike Reed", "David J. Torgerson" ], "abstract": "Recruitment remains an issue when conducting randomised controlled trials (RCTs) with a significant proportion of studies failing to reach their target sample size. Studies evaluating interventions to improve recruitment aimed specifically at recruiters to the trial are limited in number. This factorial RCT will evaluate the effectiveness of an educational intervention to trainee principal investigators and a positive reinforcement intervention via an email nudge on increasing recruitment. The targeted recruiters will be in 20 centres nationally recruiting to one large orthopaedic randomised controlled trial, WHiTE 8 COPAL. Centres will be randomised via minimisation to one of four groups. The primary outcome is recruitment rate in the first six months that a centre is actively recruiting, with data being analysed via a Poisson regression model. Results will be presented as adjusted incidence rate ratios with 95% confidence intervals. Secondary outcomes relate to the feasibility and logistics of running the interventions. We will also collect feedback regarding the educational programme set out for the trainee principal investigators. The study started in August 2018 with the anticipation of the primary objective endpoint by October 2019. The results of this study will be used to inform the design of future RCTs, particularly in orthopaedics in the UK, where the role of Trainee Principal Investigators is now a consistent one across different trials. Trial registration: 11600053, ISRCTN, 20/08/2018; SWAT 67, Northern Ireland Hub for Trials Methodology Research SWAT repository, 01/10/2017.", "keywords": [ "Trainee Principal Investigator", "TPi", "Nudge", "Swat", "Recruitment", "Education" ], "content": "Introduction\n\nRandomised controlled trials (RCTs) are considered the gold standard when evaluating the efficacy and effectiveness of health care interventions. Unfortunately, a significant number of well-designed RCTs struggle with the recruitment of participants and subsequently fail to reach their target sample size1.\n\nSeveral hypothetical and real-life studies on methods to improve participant recruitment to RCTs have been conducted with mixed results, with only a minority targeting recruiters to the trial2–5. The results of a survey of clinical trials units in the UK concluded that priorities for evaluation included training site staff, methods of communication with patients and incentivising site staff6.\n\nThe aim of this real-life study is to assess the effects of targeting healthcare professional recruiters with an educational intervention with or without positive reinforcement on participant recruitment. This study within a trial (SWAT) will test two different methods of enhancing recruitment: introducing an enhanced trainee principal investigators (TPI) package, and personalised email nudges (see Extended data7) to healthcare professionals involved in patient recruitment. The SWAT will be implemented in a large, UK, multicentre orthopaedic RCT, the WHiTE 8 COPAL trial. The interventions have both been used in current orthopaedic trials, but their effects on recruitment have previously not been investigated.\n\n\nInterventions\n\nThis will be a multicentre, 2×2 factorial RCT run between August 2018 and October 2019 with random allocation of the recruiting centre to one of four groups:\n\nGroup 1: Enhanced TPI package.\n\nGroup 2: Use of a personalised email nudge to each recruiter.\n\nGroup 3: Enhanced TPI package and use of a personalised email nudge to each recruiter.\n\nGroup 4: Usual practice (neither the enhanced TPI nor personalised email nudge).\n\nFull details of each intervention have been provided as Extended data7 and are summarised in Table 1. The consent to participate as a TPI and interventions will be implemented by the White 8 Research Fellow (author N.A.). Consent materials are also available as Extended data7.\n\nTPI, trainee principal investigator\n\n\nSample selection\n\nAs in many SWATs, a power calculation was not undertaken as the number of participating sites is fixed and driven by the needs of the host trial. All WHiTE centres planned to be recruiting to the WHiTE 8 trial will be included, except the centre in which N.A. is based. We anticipate a minimum of 20 centres being involved in recruiting. Trial interventions will only be discontinued if the host trial (WHiTE 8 Copal) is discontinued.\n\n\nRandomisation\n\nThe WHiTE centres will be randomised by minimisation on a rolling basis as sites become activated to one of the four groups to balance key baseline characteristics. Self-reported site feasibility questionnaires completed by the recruitment centres will be used to collect the information required for the minimisation. Minimisation will be based on the following factors:\n\n1. Cluster size (number of intracapsular hip fractures presenting in the previous year, cut at the median <300 or ≥300)\n\n2. High vs Low recruiting centres (<9 or ≥9 per month based on previous RCTs run within WHiTE Cohort)\n\n3. Co-recruitment to WHiTE 5 (yes/no) (Another RCT using the same patient population running at a few of the recruitment sites)\n\nThis randomisation will be performed using specialist computer software, MinimPy (Saghaei and Saghaei, 2011). This is an open trial and participating sites, the data analyst nor trial team will be blind to allocation.\n\n\nOutcomes\n\nThe primary outcome is the total number of patients recruited in the first 6 months from a site opening to recruitment to the WHiTE 8 COPAL trial.\n\nThe secondary outcomes are: conversion rate from screened population collected monthly from the central recruitment database (coordinated by the Oxford Clinical Trials Research Unit); and the time taken to implement each intervention from commencing recruitment in each centre.\n\nThe trainee’s perspective of their role will be collected through the TPI survey (available as Extended data7 at the end of the SWAT in each centre. The Research Fellow will keep a record of the time taken delivering the TPI education intervention and a log of communication for peer-support during the period of the SWAT to inform future implementation.\n\n\nEthical issues\n\nThe University of York Health Sciences Ethics Committee has approved this study within a trial. Ethics Approval ID: HSRGC/2018/266/C. Substantive protocol amendments will be sought approval through then university ethics committee.\n\n\nTrial registration\n\nThis SWAT is registered with ISRCTN (11600053) and is embedded in the WHITE 8 Copal trial (ISRCTN 15606075).\n\nThis SWAT is also registered to the SWAT repository store as part of the Northern Ireland Hub for Trials Methodology Research (SWAT 67).\n\n\nData analysis\n\nAnalysis will be conducted in STATA v15 on an intention-to-treat basis, including all sites in the group they were originally allocated to regardless of deviations based on non-compliance. Statistical significance will be assessed using logistic regression two-sided statistical tests at the 5% significance level. The trial will be reported to CONSORT guidelines, and a flow diagram will present the progression of sites through the trial.\n\nBaseline data relating to the sites (including the minimisation factors) will be summarised for the four groups as randomised and as analysed to assess whether possible loss-to-follow-up has introduced selection bias. Continuous data will be presented using descriptive statistics (e.g., mean, standard deviation, median, minimum, maximum), while categorical data will be given as counts and percentages. No formal statistical comparison of baseline data will be undertaken between the four groups.\n\nThe number of participants recruited per site will be summarised. A Poisson regression model, containing the two interventions (Enhanced TPI and Email Nudge) and the minimisation factors (cluster size, and number recruited per month will be included in their continuous form) will be undertaken. Adjusted incidence rate ratios (IRRs) and associated 95% confidence intervals (CIs) will be obtained from this model. The presence of an interaction between the two interventions will also be tested by including an interaction term in the model.\n\nFeasibility outcomes, such as the time required to run the education intervention and communication time and methods used for the peer support aspect of the intervention, will be reported descriptively.\n\nA data monitoring committee will not be used as this a trial involving recruiters and patient safety will not be affected by conducting this trial. No formal auditing of trial procedure will take place.\n\n\nDiscussion\n\nIf successful, we would like to show that these can be feasibly implemented in future RCTs with additional benefit of reaching targeted sample sizes within the planned recruitment timeline due to increased recruitment rates.\n\n\nPlans for dissemination\n\nResults of this study will be form part of a PhD thesis, published in a peer-reviewed journal, presented at conferences and be shared with recruiting centres and clinical trials units.\n\n\nData availability\n\nNo underlying data are associated with this article.\n\nOpen Science Framework: Protocol for a factorial randomised controlled trial, embedded within WHiTE 8 COPAL, of an Enhanced Trainee Principal Investigator Package and Additional Digital Nudge to increase recruitment rates. https://doi.org/10.17605/OSF.IO/FZ4JH7.\n\nThis project contains the following extended data:\n\n• Extended SWAT Protocol\n\n• Nudge email 1\n\n• NUDGE MATRIX\n\n• TrainingPackage_V1_2017-03-14 (trainee principal investigator manual)\n\n• Consent for contact\n\n• Enhanced TPi Induction Agenda ver1.0apr18\n\n• New TPI Checklist\n\n• SWAT Participation info ver3apr18\n\n• TPI Induction Presentation\n\n• TPi_Follow_up_Survey\n\nOpen Science Framework: SPIRIT checklist for article ‘Protocol for a factorial randomised controlled trial, embedded within WHiTE 8 COPAL, of an Enhanced Trainee Principal Investigator Package and Additional Digital Nudge to increase recruitment rates’. https://doi.org/10.17605/OSF.IO/FZ4JH7.\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nWalters SJ, Bonacho Dos Anjos Henriques-Cadby I, Bortolami O, et al.: Recruitment and retention of participants in randomised controlled trials: a review of trials funded and published by the United Kingdom Health Technology Assessment Programme. BMJ Open. 2017; 7(3): e015276. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTreweek S, Pitkethly M, Cook J, et al.: Strategies to improve recruitment to randomised trials. Cochrane Database Syst Rev. 2018; 2: MR000013. PubMed Abstract \| Publisher Full Text\n\nLiénard JL, Quinaux E, Fabre-Guillevin E, et al.: Impact of on-site initiation visits on patient recruitment and data quality in a randomized trial of adjuvant chemotherapy for breast cancer. Clin Trials. 2006; 3(5): 486–492. PubMed Abstract \| Publisher Full Text\n\nMonaghan H, Richens A, Colman S, et al.: A randomised trial of the effects of an additional communication strategy on recruitment into a large-scale, multi-centre trial. Contemp Clin Trials. 2007; 28(1): 1–5. PubMed Abstract \| Publisher Full Text\n\nFletcher B, Gheorghe A, Moore D, et al.: Improving the recruitment activity of clinicians in randomised controlled trials: a systematic review. BMJ Open. 2012; 2(1): e000496. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBower P, Brueton V, Gamble C, et al.: Interventions to improve recruitment and retention in clinical trials: a survey and workshop to assess current practice and future priorities. Trials. 2014; 15(1): 399. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAgni N: Protocol for a factorial randomised controlled trial, embedded within WHiTE 8 COPAL, of an Enhanced Trainee Principal Investigator Package and Additional Digital Nudge to increase recruitment rates. 2019. http://www.doi.org/10.17605/OSF.IO/FZ4JH" }	[ { "id": "52423", "date": "03 Sep 2019", "name": "Mike Clarke", "expertise": [ "Reviewer Expertise Clinical trials", "systematic reviews", "research methodology." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nStruggling or failing to achieve their target recruitment, and therefore, failing to have adequate statistical and clinical power, is a problems for many clinical trials. It contributes to research waste and slows down our ability to resolve important uncertainties in health and social care.\nThis is a well reported protocol for a SWAT (Study Within A Trial) that will be embedded in the WHiTE 8 COPAL randomised trial. The SWAT will investigate ways to increase recruitment in the clinical trial. I think it should be indexed and I wish the authors well with the conduct and analysis of the SWAT. It might be helpful to add a reference to a more general article about SWAT to help readers who would like to find out more (e.g. Education section – Studies Within A Trial (SWAT). (2012)1; or Smith V, Clarke M, Devane D, et al. (2013)2; or Treweek S, Bevan S, Bower P, et al. (2018)3).\nThe methodology study has been registered as SWAT 67 (go.qub.ac.uk/SWAT-SWAR) and is also registered as ISRCTN11600053. It will be a factorial, cluster randomised trial in which recruiters at each of the 20 sites for WHiTE 8 COPAL will be allocated to one of four groups. The two interventions in the factorial design are the introduction of an enhanced trainee principal investigators' (TPI) package and personalised email nudges, which are sufficiently different to meet the criteria for the good use of a factorial trial design. There is an adequate summary of the interventions in the body of the article. The use of minimisation to allocate the centres to one of the four groups is an appropriate strategy to help achieve balance, especially with such a relatively small number of units being randomised. Apart from my concern below, the minimisation variables seem reasonable. However, might variables 1 (cluster size) and variable 2 (monthly recruitment in previous randomised trials in WHiTE COPAL) be so closely correlated that they will serve as a single variable only? For instance, my presumption would be that the smaller centres (i.e. <300 cases) would have the lower recruitment (i.e. <9/month), and the larger centres would have the higher recruitment.\nMy concern is with the primary outcome, which is the number of patients recruited in the first six months after a centre opens. My concern is that there might be such wide variation in these numbers that, with perhaps five centres in each of the four allocation groups (or ten in each of the groups testing each of the two interventions) that underlying differences between the centres might dominate the final results. I realise that cluster size will be used in the minimisation and in the analysis but this seems to be dichotomous around a median of 300 cases in the previous year which, depending on the distribution of the cluster sizes, might be too simplistic a split. For example, if there are three sites with less than 50 cases in the previous year, or three with more than 500 cases, these might highly skew the data if, by chance, all three at the low or at the high end are randomly allocated to one of the interventions. It would be reassuring to know that the centres are not so heterogeneous or, if they are, that a finer level of adjustment than below and above 300 will be used in the analysis.\nIn summary, this is an important study. It will provide evidence relevant to other orthopedic trials, that may help future researchers to boost recruitment (if one or both interventions are effective) or to re-direct SWAT to other areas if they are not. Its relevance is also likely to extend beyond clinical trials in this setting, by contributing to the overall evidence base on interventions to boost recruitment, which have been shown to be so lacking in the Cochrane Methodology Review (reference 2 in this manuscript).\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Not applicable", "responses": [] }, { "id": "52279", "date": "05 Sep 2019", "name": "Kath Star", "expertise": [ "Reviewer Expertise Trial management and design" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting and well designed trial and the outcomes will be of interest to many involved in running clinical trials.\n\nThere are a few suggestions/areas for consideration by the authors:\n\nThe manuscript would benefit from the addition of a description of the role of a TPI should be included (as stated in section 1.3.2 of the extended protocol)\n\nHave the authors considered additional timepoints for follow up? Recruitment fatigue can often affect long-running studies and it would be interesting to determine whether the intervention has longer-lasting effects.\n\nFeasibility and resource use are an important aspect of the study and should be included in the secondary outcomes.\n\nHave the authors considered performing a formal cost analysis? It would be interesting to understand the cost/benefit of the intervention.\n\nThere are a number of acronyms referring to the studies hosting or associated with EnTraP – WHiTE, WHiTE 8, WHiTE 8 Copal, WHiTE 5 - that we found confusing and detracted from the description of the study. Section 1.3.1 of the extended protocol explains the relationship between these studies well, and it would be useful to include this for the reader to understand the context of the host trial.\n\nIn the extended protocol the study is referred to as EnTraP and it would be beneficial to include this in the article to refer to e.g. ‘the EnTraP Research Fellow’.\n\nWill education and experience of the sites staff be mentioned or evaluated when analysing the results as this can impact on recruitment.\nThe extended protocol submitted as extended data still has some comments and tracked changes the authors may want to remove before indexing.\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Partly\n\nAre the datasets clearly presented in a useable and accessible format? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1153
https://f1000research.com/articles/7-1915/v1	10 Dec 18	{ "type": "Research Article", "title": "Considerations related to vaping as a possible gateway into cigarette smoking: an analytical review", "authors": [ "Peter N. Lee", "Katharine J. Coombs", "Esther F. Afolalu", "Katharine J. Coombs", "Esther F. Afolalu" ], "abstract": "Background: Toxicant levels are much lower in e-cigarettes than cigarettes. Therefore, introducing e-cigarettes into the market seems likely to reduce smoking-related diseases (SRD). However, vaping might provide a gateway into cigarette smoking for those who otherwise would never have smoked, a concern fueled by cohort studies showing vaping predicts subsequent smoking initiation in young people. Methods: In this discussion paper, we consider various aspects of the gateway issue in youths. We provide a descriptive critical review of results from prospective studies relating to the gateway effect and the extent to which the studies considered other potential confounding variables associated with smoking initiation. We then estimate the effects of omitting a confounding variable, or misclassifying it, on the association between vaping and subsequent smoking initiation, and determine how the prevalence of smoking might be affected by any true gateway-in effects of vaping. Finally, we examine trends in e-cigarette and smoking prevalence in youths based on national surveys. Results: First, we demonstrate that although studies report that vaping significantly predicts smoking initiation following adjustment for various other predictors, the sets of predictors considered are quite incomplete. Furthermore, no study considered residual confounding arising from inaccurate measurement of predictors. More precise adjustment may substantially reduce the association. Second, we show any true gateway effect would likely affect smoking prevalence only modestly. Third, we show smoking prevalence in U.S. and U.K. youths in 2014–2016 declined somewhat faster than predicted by the preceding trend; a substantial gateway effect suggests the opposite. Finally, we argue that even if some gateway effect exists, introducing e-cigarettes still likely reduces SRDs. Conclusions: We have shown that the existence of any true gateway-in effect in youth is not yet clearly demonstrated and that the population health impact of introducing e-cigarettes is still likely to be beneficial.", "keywords": [ "Cigarettes e-cigarettes", "gateway effects" ], "content": "Introduction\n\nRecent publications made clear that, in youths, vaping (i.e., use of e-cigarettes) and cigarette smoking (subsequently referred to as “smoking”) are strongly associated. In the U.S., for example, a survey of ninth and tenth grade students in Hawaii in 2014 (Wills et al., 2017) revealed 195 ever-users of both products, 250 ever-vapers only, 37 ever-smokers only, and 820 never-users of either. From these data, the odds ratio (OR) relating ever-vaping to ever-smoking can be estimated as 17.3 (95% confidence interval (CI) 11.8–25.3). A strong association can also be seen for sixth to twelfth grade students in Texas in 2014 (Cooper et al., 2016), with an OR of 28.8 (25.0–33.1), as well as nationally in 2012 (Dutra & Glantz, 2014), with an OR of 31.9 (27.6–36.8). Similar strong relationships are reported also in Canada (Aleyan et al., 2018), France (Dautzenberg et al., 2016), Great Britain (Eastwood et al., 2015), Poland (Goniewicz et al., 2014), and Korea (Lee et al., 2014). Theoretically, this association may arise if vaping encourages smoking, if smoking encourages vaping, and/or if other factors link to use of both products.\n\nConcern that vaping may encourage subsequent smoking, the so-called “gateway-in” effect, was fuelled by a recent paper by Soneji et al. (2017). In this manuscript, the authors combined epidemiological evidence from nine U.S. cohort studies in young people that linked smoking initiation to previous vaping. Among baseline never-smokers, baseline ever-vaping strongly predicted smoking initiation within six to 18 months (OR 3.62, 95% CI 2.42–5.41) after adjusting for various predictors of initiation. Similarly, baseline past 30-day vaping predicted subsequent 30-day cigarette use (OR 4.28, 95% CI 2.52–7.27).\n\nBased on these results and those from one additional study (Hammond et al., 2017), the National Academies Press on the Public Health Consequences of E-Cigarettes (National Academies of Sciences Engineering and Medicine, 2018) recently concluded that “there is substantial evidence that e-cigarette use increases risk of ever using combustible tobacco cigarettes among youth and young adults, noting that the relevant studies had adjusted for “a wide range of covariates” and considering that it was “unlikely that confounding entirely accounts for the association because reductions in estimates of association from unadjusted to adjusted models were not consistently observed in the literature.”\n\n\nMethods\n\nThis review considers various methodological aspects of the gateway issue in youths.\n\nFirst, based on PubMed searches carried out at intervals starting in May 2017 on “ecigarettes” or “e-cigarettes” or “e-cigs” or “electronic cigarette” or “ecigarette” or “e-cigarette”, we identified papers and reviews that presented results from prospective studies of young people relating to the gateway-in effect. Further publications were also sought from reference lists of selected papers and reviews. For this review, we also considered studies that provided data on trends over time in cigarette smoking by youths in relation to e-cigarette use, and information on initiation of smoking and e-cigarette smoking by youths relevant to the gateway effect.\n\nA list of factors other than e-cigarette use that were associated with the initiation of cigarette smoking was obtained from US Surgeon General (1994) and from a separate ongoing review on determinants of smoking initiation which aims to present meta-analyses of associations between initiation of cigarette smoking and the various other factors. This identified references using an Embase search in April 2017 with the following search string:\n\n'smoking'/de OR 'smoking' AND ('initiation'/de OR 'initiation') AND ('prevalence'/de OR 'prevalence' OR 'incidence'/de OR 'incidence' OR 'proportion' OR 'age'/de OR 'age' OR 'dual use' OR 'combined use' OR 'psychosocial'/de OR 'psychosocial' OR 'beliefs'/de OR 'beliefs' OR 'attitudes'/de OR 'attitudes' OR 'perceptions'/de OR 'perceptions' OR 'opinions' OR 'acceptance'/de OR 'acceptance' OR 'predictors'/de OR 'predictors' OR 'friend'/de OR 'friend' OR 'family'/de OR 'family' OR 'parent'/de OR 'parent' OR 'propensity score'/de OR 'propensity score') AND [humans]/lim AND [english]/lim AND ([embase]/lim OR [medline]/lim)\n\nFor the current paper, we only use the results from this search to list those factors shown in one or more studies to strongly associate with initiation of smoking, and to consider the extent to which the identified gateway studies take these factors into account.\n\nIn addition, separate sheets of an Excel Program (Lee, 2018) were developed to:\n\n(a) estimate the effects of omission of a confounding variable, or misclassification of it, on the association between vaping and subsequent initiation of smoking, and to\n\n(b) determine how the prevalence of smoking might be affected by way of any true gateway-in effects of vaping.\n\nFull details are given in, respectively, Additional Files 2 and 3 (Lee, 2018). Both sheets are used to provide illustrative examples of how the effects depend on the parameter values assumed.\n\nIn the first sheet, the use may determine how the observed gateway-in effect depends on the following parameters:\n\nThe proportion of vapers in never smokers (P1)\n\nThe proportion with a particular smoking predictor present in never smokers (P2)\n\nThe concordance (odds ratio) between vaping and the smoking predictor (K)\n\nThe probability of initiating smoking during follow-up in those who neither smoke nor have the predictor (PA)\n\nThe odds ratio for initiating smoking during follow-up associated with vaping (the true gateway-in effect – GE)\n\nThe odds ratio for initiating smoking during follow-up associated with the smoking predictor (GP)\n\nThe proportion with the predictor who are misclassified as not having it (M1), and\n\nThe proportion without the predictor who are misclassified as having it (M2)\n\nP1, P2 and K define the baseline population in which N never smokers are subdivided in a 2 x 2 table according to presence or absence of e-cigarettes and of the smoking predictor, while PA, GE and GP determine the distribution of the population after a single follow-up period. M1 and M2 are misclassification rates of the predictor.\n\nIn the second sheet, the program considers individuals, divided into five equal strata, who initially are all never users of either cigarettes or e-cigarettes. The probability of initiating smoking or vaping increases progressively over the strata, the strata being intended to represent sets of individuals with an increasing susceptibility to tobacco. Over five time intervals, the individuals may transfer to four other groups: current vapers only, current smokers only, current dual users and former users. Users may modify the values of seven parameters:\n\nThe initiation rate of vaping in the first stratum (P1)\n\nThe relative odds of initiation for successive strata (R1)\n\nThe relative odds of initiation with smoking compared to vaping (R2)\n\nThe relative odds of quitting compared to initiation (R3)\n\nThe relative odds of re-initiation compared to initiation (R4)\n\nThe relative odds of initiating smoking for vapers compared to tobacco never-users (“gateway-in” effect)\n\nThe relative odds of quitting smoking for dual users compared to smokers only (“gateway-out” effect)\n\nLastly, the review also includes an analysis of trends in e-cigarette and smoking prevalence in youths based on national surveys which provided annual information over a period of at least 10 years. We identified four surveys from the U.S., the Youth Risk Behavior Survey (YRBS: available from https://www.cdc.gov/healthyyouth/data/yrbs/index.htm), the National Youth Tobacco Survey (NYTS: available from https://www.cdc.gov/tobacco/data_statistics/surveys/nyts/index.htm), the Monitoring the Future study (MTF: available from https://www.icpsr.umich.edu/icpsrweb/NAHDAP/series/35), and the National Survey on Drug Use and Health (NSDUH: available from https://www.samhsa.gov/data/data-we-collect/nsduh-national-survey-drug-use-and-health). We also identified two surveys from the UK, the Smokefree Youth Survey in Great Britain (SYSGB: available from http://ash.org.uk/download/use-of-electronic-cigarettes-among-children-in-great-britain/) and the Office for National Statistics (ONS: available from https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/drugusealcoholandsmoking/datasets/ecigaretteuseingreatbritain). All six databases were accessed on 16 Oct 2017.\n\nBased on linear regression of loge (p/(1−p)) a year, where p is prevalence, observed prevalences in the years 2014–2016 (a period where the advent of e-cigarettes might have had a measurable effect on smoking prevalence if an important gateway-in effect existed) was compared with those expected based on the trend in the previous years for which results were available.\n\n\nResults\n\nFrom the literature search, we identified a key systematic review and meta-analysis by Soneji et al. (2017) on the association between initial use of e-cigarettes and subsequent cigarette smoking among adolescents and young adult. Table 1 provides details of the nine U.S. studies analyzed by Soneji et al. (2017). A total of five studies, conducted with participants with a maximum age of 24–30 years old, collected information using internet-based surveys. The four other studies, involving participants aged less than 20 years old were based on self-completed questionnaires. The follow-up period was 6 months in one study (Hornik et al., 2016), 1 year in five studies (Leventhal et al., 2015; Primack et al., 2015; Spindle et al., 2017; Unger et al., 2016; Wills et al., 2017), and 13 to 18 months in the other three studies (Barrington-Trimis et al., 2016; Miech et al., 2017; Primack et al., 2016). Most studies concerned ever- and never-use, and two studies considered current and noncurrent use. Each study found that baseline vaping significantly predicted subsequent smoking, regardless of covariate adjustment. Adjusted ORs were lower than unadjusted ORs in six studies, particularly in two (Hornik et al., 2016; Leventhal et al., 2015). However, two studies (Primack et al., 2015; Primack et al., 2016) showed a moderately increased OR after adjustment.\n\naTwo studies (Hornik et al., 2016; Primack et al., 2016) were reported only as abstracts, but fuller details were supplied to Soneji et al. for their meta-analyses. bOf never- (or noncurrent) smokers at baseline. cA = Odds of smoking initiation, among never-smokers at baseline, for ever- compared with never-vapers at baseline. B = Odds of smoking at baseline, among noncurrent smokers at baseline, for current compared with noncurrent vapers at baseline. dORs for U.S. studies as given by Soneji et al. ORs for U.K. studies come from the source, except that the unadjusted estimates for Best et al. and Loukas et al. were estimated from data given.\n\nTable 1 also includes results from six later studies cited in a recent review (Glasser et al., 2018): two from the U.S. (Loukas et al., 2018; Watkins et al., 2018), two from the U.K. (Best et al., 2018; Conner et al., 2018), and one each from Canada (Hammond et al., 2017) and Mexico (Loukas et al., 2018). These results showed an association that reduced but remained significant after adjustment. Two further studies could not be included in Table 1, as they did not provide comparable results. One was a study on young adults in the Chicago area (Selya et al., 2018) that used path analysis and concluded that “E-cigarette use was not significantly associated with later conventional smoking…” The other was a study in the Netherlands (Treur et al., 2018) that reported a strong association after covariate adjustment but did not present unadjusted results for comparison.\n\nThere were an additional 15 publications (Amato et al., 2016; Ambrose, 2017; Barrington-Trimis et al., 2015; Bold et al., 2016; de Lacy et al., 2017; Doran et al., 2017; Etter & Bullen, 2014; Hanewinkel & Isensee, 2015; Huh & Leventhal, 2016; Kaufman et al., 2015; Leventhal et al., 2016; Loukas et al., 2015; Sutfin et al., 2015; Westling et al., 2017; Zhong et al., 2016) identified in our searches as of possible relevance based on the abstract. However, none of these provided useful data for various reasons. These included studies which: were conducted in adults; predicted e-cigarette use rather than cigarette smoking; did not present results in a form allowing the gateway effect to be estimated; were superseded by later publications; concerned intent to smoke cigarettes and not actual smoking; or even not considering e-cigarettes at all.\n\nInitiation of smoking is long-established to be associated with various sociodemographic, environmental, and behavioral factors. Those identified by an authoritative source over 20 years ago (US Surgeon General, 1994) include low education level (Gritz et al., 1998; Sargent et al., 1997), internalizing/externalizing disorders (de Leon et al., 2002; Ernst et al., 2010; Rohde et al., 2003), outcome expectancies (Barrington-Trimis et al., 2015; Simons-Morton et al., 1999), susceptibility to smoking (Huang et al., 2005; Jackson, 1998), conduct problems (Dalton et al., 2003; Scal et al., 2003), substance use (Reed et al., 2007; Scal et al., 2003), risk-taking behavior (Coogan et al., 1998; Dalton et al., 2003), poor school performance (O'Connor et al., 2003; Sargent et al., 1997), anxiety (Coogan et al., 1998; Scal et al., 2003), household smoking (Picotte et al., 2006; Sargent et al., 1997), peer smoking (Barrington-Trimis et al., 2015; Picotte et al., 2006), peer attitudes to smoking (Barrington-Trimis et al., 2015; Daly et al., 1993), low self-esteem (Dalton et al., 2003; Weiss et al., 2006), and other tobacco product use (Barrington-Trimis et al., 2016; Jordan et al., 2014). Currently available evidence reveals that many of these predictors consistently show a strong association with initiation, with reported ORs often exceeding 10. While these predictors are clearly not all independent, their number illustrates the difficulty in ensuring that gateway studies consider an adequate list. To avoid unfeasibly long questionnaires or huge studies, further work is needed to define an agreed minimum list of factors.\n\nFor the 15 studies included in Table 1, Additional File 1 (Lee, 2018) describes in detail how these risk factors were taken into account. While standard demographics were generally considered, and smoking susceptibility, substance use, risk-taking behavior, other tobacco use, parental education, family smoking, and peer smoking were considered in at least five studies, many other relevant factors were rarely considered. These include internalizing/externalizing disorders, outcome expectancies, school performance, anxiety, parental smoking, and peer attitudes to smoking. The studies vary in the number of factors accounted for: some (Conner et al., 2018; Leventhal et al., 2015; Watkins et al., 2018; Wills et al., 2017) considered more than 10 factors, others (Miech et al., 2017; Unger et al., 2016) only five. As shown in Additional File 1 (Lee, 2018), the questions used to assess these factors also varied considerably between studies.\n\nUnfortunately, no study reported which factors contributed most to their adjustment. Thus, for example, Leventhal et al. (2015) adjusted for the most factors and found that the unadjusted OR of 7.78 (6.15–9.84) reduced dramatically after adjustment to 1.75 (1.10–2.78); however, the authors did not clarify which factors mainly contributed to this reduction.\n\nWhat effect might failure to adjust for a relevant predictor of smoking have on the estimated gateway-in effect observed between vaping and smoking? To answer to this question, we consider (using the first sheet of the Excel program we developed) a hypothetical baseline population of N never-smokers, of which P1 vape, and P2 have the smoking predictor. P1 and P2 are correlated, with a concordance ratio of K. We assume that, during follow up, those who neither have smoked nor have the predictor have a probability of initiating smoking of PA and that the odds of smoking initiation are independently increased by GE for vapers and by GP in those with the smoking predictor.\n\nThe illustrative results in Table 2, supported by mathematical detail in Additional File 2 (Lee, 2018), demonstrate the problem. In the basic Situation 1, we set P1 = 0.2, P2 = 0.2, K = 5, PA = 0.05, GE = 1, and GP = 4. Instead of observing the true gateway effect (GE) of 1, we observe a spurious gateway effect of 1.633 due to the uncontrolled confounding. In Situations 2 through 6, GE remains at 1, but other parameters are varied. There is a clear tendency for the confounding effect to increase with increasing K (Situation 2) and GP (Situation 3), but varying PA (Situation 4), P1 (Situation 5), or P2 (Situation 6) have less effect. Increasing GE, with the other parameters fixed (Situation 7), increases the observed gateway effect, but the proportional increase in the OR is reduced.\n\nWhile confounding effects are well understood by epidemiologists, which is why the authors of the gateway-in papers made some adjustments, “residual confounding,” arising from inaccurately determining confounding variables, is rarely considered. Many statisticians have highlighted this problem of residual confounding. Almost 40 years ago, Greenland (1980) noted that “misclassification of a confounder” leads to “partial loss of ability to control confounding,” while Tzonou et al. (1986) later noted that “even misclassification rates as low as 10% can prevent adequate control of confounding.” Other publications (Ahlbom & Steineck, 1992; Fewell et al., 2007; Greenland & Robins, 1985; Phillips & Smith, 1994; Savitz & Barón, 1989) show that if X is an inaccurately measured true cause of disease, and Y, precisely measured, is not a cause but is correlated with X, one may conclude incorrectly that Y, not X, is the cause. None of the gateway papers cited in Table 1, nor Soneji et al. (2017), refer to residual confounding as a potential source of bias, although all except two (Primack et al., 2015; Primack et al., 2016) mention possible bias from incomplete adjustment.\n\nTable 3 (see also Additional File 2 (Lee, 2018)) gives illustrative examples of the effect of residual confounding. Situations 1 to 4 concern misclassification occurring in each direction corresponding to situations 1, 2(a), 3(a), and 7(a) in Table 2, where no misclassification is assumed, whereas in Table 2, the smoking predictor is assumed to be measured accurately, and adjustment for it would correct the observed gateway-in effect back to its true value (GE); this is not true when misclassification is present. Thus, in Situation 1, where GE is set at 1, adjustment for the misclassified predictor does not fully correct for the confounding. Adjustment is useless where the misclassification rate is 50% with the observed gateway-in effect staying at its unadjusted value of 1.633. With a 10% misclassification rate, the value reduces to 1.281 so that about half (281/633 = 44.4%) of the spurious increase in the OR remains. As misclassification rates reduce, the adjusted rate approaches its true value.\n\nNotes: A misclassification rate of x% in both directions (considered in situations 1 to 4) implies that x% of those who have the factor that predicts smoking (e.g. risk takers) are wrongly classified as not having it, and x% of those who do not have the factor are wrongly classified as having it. Situations 5 and 6 consider misclassification in one direction.\n\nAll models have P1 = P2 = 0.2 and PA = 0.5.\n\nThe pattern is similar in Situations 2 and 3, where GE remains at 1, but K and RP are varied. Again, all of the bias remains with 50% misclassification, and almost half remains with 10% misclassification. This is also observed in Situation 4, where GE = 2. Situations 5 and 6 are similar to Situation 1, except misclassification is assumed to be unidirectional. The bias is similar to that observed in Situation 1, where only false positives occur (Situation 5), but less where only false negatives are present (Situation 6). Poor specificity is more important than poor sensitivity as a source of bias.\n\nIt is important to understand how vaping might affect the prevalence of cigarette smoking. Not only might there be “gateway-in” effects, with vaping increasing the probability of subsequent smoking, but there might also be also “gateway-out” effects, with vaping by smokers increasing their probability of smoking cessation. In an attempt to gain some understanding, we used the second sheet of the Excel program we developed (see also Additional File 3 (Lee, 2018)). The program considers 2,000 individuals in each of five strata that represent groups with varying smoking susceptibility. All individuals start as never-users of either product, and over time, may switch to become current users of either product only, current dual users, or former users.\n\nWhere no gateway effects apply, the user may vary the values of P1, R1, R2, R3 and R4 as defined in the methods section. These parameters are then used to update tobacco use status over time. Where gateway effects apply, the user may define G1, the gateway-in effect, and G2, the gateway-out effect.\n\nTable 4 presents illustrative examples where no gateway-in effects apply. The five parameters are varied in turn, with the other parameters held constant. Increasing e-cigarette initiation rates by increasing P1 (Block 1) reduces never-users, with a compensating increase in dual users and in concordance. In Block 1, it is assumed that there is no quitting or re-initiation (R3 = R4 = 0), the relative odds of initiation for successive strata (R1) are set at 5, and initiation rates are assumed the same for both tobacco products (R2 = 1). Increasing R1 (Block 2), which increases the between stratum variation in smoking susceptibility, markedly increases the concordance ratio. The other Blocks keep P1 at 0.0002 and R1 at 5. Increasing R2 (Block 3) increases the frequency of smoking relative to vaping but also to dual use and the concordance ratio. Introducing quitting factor (Block 4) reduces smokers, but concordance is less affected. Furthermore, introduction of re-initiation factor (Block 5) has little effect; in fact, the chance of someone initiating, quitting, then re-initiating in this period is small.\n\nAll models assume no gateway effects.\n\nP1, initiation rate of vaping in stratum 1; R1, relative odds of initiation for successive strata; R2, relative odds of initiation for smoking compared to vaping; R3, relative odds of quitting compared to initiation; R4, relative odds of re-initiation compared to initiation.\n\nTable 5–Table 7 show how the relative odds of smoking are affected by gateway-in effects only, gateway-out effects only, or both, respectively. All these results set P1 = 0.0002, R1 = 5, and R4 = 0. With no gateway effects, there are 11.28% cigarette smokers at the end of follow up: 7.43% smoking cigarettes only, and 3.85% dual users.\n\nG1, is the gateway-in effect; R2, the relative odds of initiation for smoking compared to vaping. The assumed values of the other parameters are P1 = 0.0002, R1 = 5, R3 = 0.2, R4 = 0, and G2 = 1. See text for the full definitions of these other parameters.\n\nOdds are expressed relative to the no gateway situation, where B1 and G2 = 1.\n\nG1 is the gateway-out effect. R2 is the relative odds of initiation for smoking compared to vaping. R3 is the relative odds of quitting compared to initiation. The assumed values of the other parameters are P1 = 0.0002, R1 = 5, R4 = 0, and G1 = 1. See text for the definitions of these other parameters.\n\nG1 is the gateway-in effect. G2 is the gateway-out effect. R2 is the relative odds of initiation for smoking compared to vaping. R3 is the relative odds of quitting compared to initiation. The assumed values of the other parameters are P1 = 0.0002, R1 = 5, and R4 = 0. See text for the definitions of these other parameters.\n\nAs the gateway-in effect (G1) increases to 5 (Table 5), the percentage of cigarette smokers rises to 13.96%, giving an OR for smoking of 1.28 compared with G1 = 1. The increase in the percentage of smokers is much less than the increase in G1; in fact, there are typically far more never-smokers who initiate smoking than e-cigarette only users. Table 5 also shows that the gateway-in effect becomes less important as the relative initiation rate of cigarettes compared with that of e-cigarettes (R2) increases.\n\nAs the gateway-out effect (G2) increases from 1 to 5 (Table 6), the percentage of smokers declines. The effect is less than for increasing G1, and it is in the opposite direction. This effect is increased as R3, the relative frequency of quitting to initiation, is increased. Table 6 also shows that the gateway-out effect becomes more important as R2 increases.\n\nIn Table 7, both gateway effects are varied, with the relative odds of smoking shown for combinations of G1 and G2 = 1, 2, or 5 and for varying values of R2 and R3. The highest relative odds of smoking (1.497) are seen for the highest values of G1 and R2 and the lowest values of G2 and R3, while the lowest odds of smoking (0.890) are seen in the reverse situation.\n\nThe effects sometimes approximately cancel out. Given R3 < 1 (and it seems unlikely that quit rates would exceed initiation rates), gateway-in effects tend to exceed gateway-out effects where initiation rates are similar for vaping and smoking. However, where initiation rates for smoking are substantially higher (R2 = 4), similar gateway-in and gateway-out effects also approximately cancel out. Here, with relatively more smokers to quit, gateway-out effects are more relevant.\n\nWhile the above results are illustrative, one can derive five main general conclusions from them:\n\n1. Any increase in smoking initiation due to vaping is proportionately much smaller than any increase in smoking prevalence.\n\n2. Gateway-in effects increase if initiation with vaping is relatively more than initiation with smoking.\n\n3. Increasing the gateway-out parameter G2 decreases smoking prevalence, but less than that from a similar increase in the gateway-in parameter G1.\n\n4. Gateway-out effects increase as R2 and R3 increase, where there are more smokers who can quit.\n\n5. With both gateway effects present, the overall effect on smoking prevalence may be in either direction.\n\nSmoking prevalence in U.S. youths had declined substantially even before vaping became popular. For example, the YRBS reported a decline in past 30-day smoking from 34.8% in 1995 to 15.7% in 2013. Has introducing e-cigarettes halted or even reversed this decline?\n\nThe NYTS is the only U.S. youth survey providing trend data on e-cigarette use over a reasonably long time period. In that survey, the prevalence of past 30-day vaping in each year from 2011–2015 was 1.0%, 2.0%, 2.9%, 9.2%, and 11.1%, respectively, for students in grades 6–12. In 2011–2013, vaping seems too rare to allow for assessment of any effect on smoking prevalence. To detect any effect, it seems more appropriate to compare the prevalence in 2014–2016 with that predicted from pre-existing trends.\n\nTable 8 presents results from this comparison based on the U.S. NYTS, YRBS, MTF, and NSDUH surveys. All estimates are for past 30-day smoking. In all the surveys, smoking prevalence in 2014–2016 was less than predicted from the underlying trend, contrary to expectations of a definitive gateway-in effect being present. The tendency for the decline in prevalence to accelerate over 2014–2016 is evident regardless of sex or age.\n\naPrevious years used for estimation of trend: NYTS: 2004, 2006, 2009, 2011-2013; MTF, ONS: 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013; YRBS: 1995, 1997, 1999, 2001, 2003, 2005, 2007, 2009, 2011, 2013; NSDUH: 2009, 2010, 2011, 2012, 2013. bMTF, Monitoring the Future; NSDUH, National Survey on Drug Use and Health; NYTS, National Youth Tobacco Survey; ONS, Office for National Statistics; YRBS, Youth Risk Behavior Survey.\n\nThe Smokefree Youth Survey in Great Britain also reported increasing vaping, with percentages of 4%, 6%, 11%, and 10% each year from 2013–2016, respectively, for 11–18-year-olds. As shown in Table 8, ONS data (for an older age group) also show no tendency for a rise in cigarette smoking given increasing vaping. Here, very small annual declines (about 0.3% in males and 0.7% in females) over 2006–2015 were followed by a much larger decline of 7% between 2015 and 2016.\n\nThese analyses, though clearly limited, do not suggest that introducing e-cigarettes has adversely affected smoking prevalence trends. If there were any gateway effect, it would be clearly outweighed by other issues, such as vaping providing an alternative to cigarettes for tobacco users, or changes in attitudes to smoking resulting from anti-tobacco prevention measures.\n\nAdditional evidence highlighted the improbability of any substantial gateway-in effect of vaping on smoking prevalence. Based on the U.S. PATH study, Pearson et al. (2018) recently reported that only a small percentage of nonsmokers would be interested in using a hypothetical modified risk tobacco product. While these analyses were based on adults, it was possible to confirm this finding from the publicly available data files for 18–24-year-olds. Thus, the proportion “very or somewhat likely” to use such a product was much lower in those who had never used tobacco (57/1745 = 3.3%) than in those who had ever done so (2375/7282 = 32.7%). This difference was similar in both sexes. Those aged 12–17 years old were also asked in the PATH study if they had seen a tobacco product that claims to be safer than other tobacco products in the past 12 months and their likelihood of using such a product in the next 30 days. Again, the proportion answering “very or somewhat likely” was much lower in tobacco never-users (174/4673 = 3.72%) than in ever-users (264/1565 = 16.87%).\n\nBased on the Canadian COMPASS study, Aleyan et al. (2018) compared rates of smoking initiation over a two-year follow-up period among a sample of 9,501 students between grades 9 and 11 (cigarette never-smokers), classified at baseline into four groups by current e-cigarette use and susceptibility to smoke and assessed using a three-item validated measure. Though the data in Figure 1 of that paper showed that in both unsusceptible and susceptible never-smokers, rates of smoking initiation were higher in current than noncurrent e-cigarette users, only an estimated 33/2646 = 1.2% of those who had tried smoking during the follow-up period were unsusceptible current e-cigarette users. These results strongly indicate that even when there is some gateway-in effect, any resulting increase in smoking prevalence would be quite small.\n\nUsers of alternative tobacco products can certainly have similar nicotine exposure to that of cigarette smokers but potentially much lower risk of smoking-related diseases (SRD), as research has suggested, for example, from the experience and prevalence of Swedish snuff (“snus”) use (Agewall et al., 2002; Bolinder et al., 1997a; Bolinder et al., 1997b; Lee, 2011; Lee, 2013). As vaping presents significantly reduced exposure to toxicants and harmful and potentially harmful constituents compared with smoking (National Academies of Sciences Engineering and Medicine, 2018), it would be expected that any harmful effects would be much lower. This hypothesis was subsequently endorsed by an expert group in 2014 (Nutt et al., 2014).\n\nThis hypothesis is also further supported by various theoretical beneficial and adverse effects of vaping (see Table 9). The first major benefit of vaping (B1) concerns individuals who, in the absence of e-cigarettes, would have initiated smoking but instead take up vaping. They should have a much lower risk of SRDs than if they had smoked, given, for example, that the expert panel (Nutt et al., 2014) estimated that relative to cigarettes, e-cigarettes likely pose a reduction in harm of up to 95%. This likely reduction in risk and harm also applies to benefit B2, where smokers who would otherwise have continued to smoke instead switch to vaping. While the benefits would take longer to emerge, they should be a considerable proportion relative to quitting smoking. The health benefit would even be greater for established smokers, where vaping helps them quit (B3).\n\nAmong the three adverse vaping effects listed in Table 9, smokers who are vaping in addition to their usual cigarette consumption (A3) is probably implausible, because most dual users are likely to control their nicotine intake and actually reduce their cigarette consumption, partially replacing cigarettes with e-cigarettes (Berry et al., 2018; McNeill et al., 2014; McRobbie et al., 2014). If so, the effect should be beneficial rather than adverse. The adverse effect for smokers switching to vaping instead of quitting smoking (A2) is not ideal but would still confer some benefits, given what is known regarding the reduced harm and toxicity of e-cigarettes.\n\nClearly, the greatest adverse effect arises for individuals who otherwise would not have smoked but are encouraged by vaping to do so (A1). However, considering the overall population health impact of this gateway-in effect, two points require emphasis. First, this is only likely to affect young people. Older people who decided not to start tobacco product use seem unlikely to start vaping, and even if they did so, it would probably be because they believe vaping to be much safer than the smoking they have already rejected (Majeed et al., 2017; Popova et al., 2018; Xu et al., 2016; Zhu et al., 2013). Second, the arguments made earlier suggest that the number of new smokers originating from any true gateway-in effect should be quite low. Any adverse effects resulting from gateway effects in young people are likely to be outweighed in the general population by beneficial effect B1, as evidenced by the substantial numbers who have vaped but not smoked. The reduction in risk of SRDs for such individuals is likely to be almost as great as any increase resulting from a gateway-in effect.\n\nWhile various authors (Hill & Camacho, 2017; Levy et al., 2018; Nutt et al., 2014) estimate that the introduction of e-cigarettes will have a beneficial population health impact, a recent publication Soneji et al. (2018) argues the opposite, presenting analyses estimating that “e-cigarette use in 2014 would lead to 1,510,000 years of life lost in the U.S. population.” The study reported that the large adverse effects from an estimated 168,000 cigarette never-smokers between 12 and 29 years old initiating cigarette smoking in 2015 and going on to become daily smokers would heavily outweigh the very small beneficial effects from an additional 2,070 current cigarette smoking adults aged 25–69 quitting smoking in 2015 and remaining abstinent for seven or more years.\n\nThese analyses have various weaknesses. First, they assume that there is no reduction in risk of SRDs for cigarette smokers who become dual users but do not quit cigarettes. Dual users are likely to reduce cigarette consumption (Brose et al., 2015; Etter & Bullen, 2014; Farsalinos et al., 2016; Farsalinos et al., 2017; McRobbie et al., 2014) and hence their risk of SRD. Second, their analysis showing large adverse effects in youths relies heavily on the estimate of the gateway effect given by Soneji et al. (2017), which may largely result from confounding. Third, their analysis showing very small benefits from increased smoking quit rates in adults depended on an estimate derived from a prior review and meta-analysis by Kalkhoran & Glantz (2016) that estimated the OR of quitting among smokers with an interest in quitting. The estimate, and indeed the whole meta-analysis, has since drawn strong criticism from experts in the field for being inaccurate and misleading (West et al., 2016).\n\n\nDiscussion\n\nOur analyses yield four major conclusions:\n\n1. While the consistently increased adjusted gateway-in effect estimates appear compelling (Table 1), they may be severely biased by ignoring established predictors of smoking initiation and residual confounding – a true causal effect remains to be demonstrated.\n\n2. If a true gateway effect were to exist, it will probably have little effect on smoking prevalence.\n\n3. No available evidence exists that increasing e-cigarette use has slowed the decline in smoking prevalence; indeed, the decline appears to have accelerated.\n\n4. Finally, introducing e-cigarettes should lead to a reduced risk of SRD.\n\nThough some publications have used the evidence for gateway effects in youth to advocate for regulations to curb vaping in the general population (Miech et al., 2017; Primack et al., 2015; Soneji et al., 2017), others suggest that this over-interprets the evidence. Kozlowski & Warner (2017) point out the need “to better understand and assess confounding variables” and consider that the prospective studies merely “support that a minority of the relatively small number of e-cigarette triers—who haven’t also been experimenting with other tobacco products already—will go on to some experimentation with cigarettes.” Also, Levy et al. (2018) present calculations showing that “a strategy of replacing cigarette smoking with vaping would yield substantial life year gains, even under pessimistic assumptions [. . .].” Our argument that the gateway-in estimates are subject to relevant uncontrolled confounding is also supported by publications showing that vaping is associated with many factors associated with smoking (Barrington-Trimis et al., 2015; Hanewinkel & Isensee, 2015; Temple et al., 2017). Unlike the implicit conclusion of the National Academies report (National Academies of Sciences Engineering and Medicine, 2018), which found that control of confounding in the gateway-in studies has been adequate, our conclusions would suggest otherwise. Our calculations indicating a lack of effect of e-cigarettes on smoking trends are further supported by Dutra & Glantz (2017), who concluded that “the introduction of e-cigarettes was not associated with a change in the linear decline in cigarette smoking among youth.” While studies (McNeill et al., 2014) have argued against the likely population benefit of e-cigarettes due to their pervasive use among youths and young adults, as we have discussed, these analyses often fail to fully consider the limitations and weaknesses of the current evidence of the associated probabilities of e-cigarette use and cigarette smoking initiation in youths.\n\nNevertheless, some limitations of our work and of the available evidence must be noted. First, many risk factors for smoking initiation are correlated, and had even more been adjusted for, this might not have significantly changed the adjusted effect estimates. Second, while we suggest that residual confounding may be important, we have not considered data on how inaccurately specific risk factors are measured. Third, we did not consider evidence (Doran et al., 2017; Leventhal et al., 2016) that vaping may be associated with heavier smoking, although this factor may also be subject to confounding problems. Fourth, our evidence on trends is based on data where the prevalence of vaping is little more than 10%. Where vaping is infrequent, any true gateway-in effect will only slightly increase smoking prevalence, as cigarette never-smokers who have not vaped will far outweigh those who have. It is thus important to continue to monitor smoking and e-cigarette use trends for further conclusive insights.\n\n\nConclusions\n\nThe existence of any true gateway-in effect has not been clearly demonstrated. Even if it does exist, and subsequently, some individuals who otherwise would not have done so start to smoke cigarettes, any effect on smoking prevalence will be quite small, and any overall population health impact of introducing e-cigarettes still likely be beneficial.\n\n\nData availability\n\nAdditional File 1. Confounding factors considered by studies of vaping as a possible gateway to smoking. The file lists those factors with published evidence of a relationship with smoking and gives those factors not considered in any of the 15 studies considered in Table 1.\n\nAdditional File 2. Estimating the effects of omission of a confounding variable, or misclassification of it, on the association between vaping and subsequent initiation of smoking. The file presents details of the methodology used.\n\nAdditional File 3. How might the prevalence of smoking be affected by any true gateway effects of vaping? The file presents details of the methodology used.\n\nGateway calcs. The file is the Excel program used to produce the results presented in Additional Files 2 and 3.\n\nAll data are available on OSF, DOI: https://doi.org/10.17605/OSF.IO/Z3ST5 (Lee, 2018).\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "Author contributions\n\n\n\nPNL conceived the paper, developed the analyses reported in Additional Files 2 and 3, and drafted the manuscript. KJC discussed the paper with PNL, searched for relevant literature, checked tables, and commented on drafts. EFA was responsible for the work described in Additional File 1, collation of data used in Table 8, commented on drafts, and contributed to the writing of the manuscript.\n\n\nGrant information\n\nThis work was supported by PMI R&D, Philip Morris Products S.A., Quai Jeanrenaud 5, CH-2000 Neuchâtel, Switzerland.\n\n\nAcknowledgements\n\nWe thank Mrs. Y. Cooper and Mrs. D. Morris for typing various drafts of this paper and assembling relevant literature. We also thank colleagues for helpful comments and Philip Morris International for financial support.\n\n\nReferences\n\nAgewall S, Persson B, Lindstedt G, et al.: Smoking and use of smokeless tobacco in treated hypertensive men at high coronary risk: utility of urinary cotinine determination. Br J Biomed Sci. 2002; 59(3): 145–149. PubMed Abstract \| Publisher Full Text\n\nAhlbom A, Steineck G: Aspects of misclassification of confounding factors. Am J Ind Med. 1992; 21(1): 107–112. PubMed Abstract \| Publisher Full Text\n\nAleyan S, Cole A, Qian W, et al.: Risky business: a longitudinal study examining cigarette smoking initiation among susceptible and non-susceptible e-cigarette users in Canada. BMJ Open. 2018; 8(5): e021080. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAmato MS, Boyle RG, Levy D: E-cigarette use 1 year later in a population-based prospective cohort. Tob Control. 2016; 26(e2): e92–e96. PubMed Abstract \| Publisher Full Text\n\nAmbrose B: E-cigarette use transitions: a case study from Waves 1 & 2 of the PATH study. In: Proceedings from 2017 SRNT Annual Meeting. Pre-Conference Workshop 6 - FDA's Population Health Standard: Balancing the Risks and Benefits in Regulatory Decision-making, Florence, Italy, 2017; 29. Reference Source\n\nBarrington-Trimis JL, Berhane K, Unger JB, et al.: Psychosocial Factors Associated With Adolescent Electronic Cigarette and Cigarette Use. Pediatrics. 2015; 136(2): 308–17. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBarrington-Trimis JL, Urman R, Berhane K, et al.: E-Cigarettes and Future Cigarette Use. Pediatrics. 2016; 138(1): pii: e20160379. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBerry KM, Reynolds LM, Collins JM, et al.: E-cigarette initiation and associated changes in smoking cessation and reduction: the Population Assessment of Tobacco and Health Study, 2013-2015. Tob Control. 2018; pii: tobaccocontrol-2017-054108. PubMed Abstract \| Publisher Full Text\n\nBest C, Haseen F, Currie D, et al.: Relationship between trying an electronic cigarette and subsequent cigarette experimentation in Scottish adolescents: a cohort study. Tob Control. 2018; 27(4): 373–378, pii: tobaccocontrol-2017-053691. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBold KW, Kong G, Cavallo DA, et al.: Reasons for Trying E-cigarettes and Risk of Continued Use. Pediatrics. 2016; 138(3): pii: e20160895. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBolinder G, Norén A, de Faire U, et al.: Smokeless tobacco use and atherosclerosis: an ultrasonographic investigation of carotid intima media thickness in healthy middle-aged men. Atherosclerosis. 1997a; 132(1): 95–103. PubMed Abstract \| Publisher Full Text\n\nBolinder G, Norén A, Wahren J, et al.: Long-term use of smokeless tobacco and physical performance in middle-aged men. Eur J Clin Invest. 1997b; 27(5): 427–433. PubMed Abstract \| Publisher Full Text\n\nBrose LS, Hitchman SC, Brown J, et al.: Is the use of electronic cigarettes while smoking associated with smoking cessation attempts, cessation and reduced cigarette consumption? A survey with a 1-year follow-up. Addiction. 2015; 110(7): 1160–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nConner M, Grogan S, Simms-Ellis R, et al.: Do electronic cigarettes increase cigarette smoking in UK adolescents? Evidence from a 12-month prospective study. Tob Control. 2018; 27(4): 365–372, pii: tobaccocontrol-2016-053539. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCoogan PF, Adams M, Geller AC, et al.: Factors associated with smoking among children and adolescents in Connecticut. Am J Prev Med. 1998; 15(1): 17–24. PubMed Abstract \| Publisher Full Text\n\nCooper M, Case KR, Loukas A, et al.: E-cigarette Dual Users, Exclusive Users and Perceptions of Tobacco Products. Am J Health Behav. 2016; 40(1): 108–16. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDalton MA, Sargent JD, Beach ML, et al.: Effect of viewing smoking in movies on adolescent smoking initiation: a cohort study. Lancet. 2003; 362(9380): 281–5. PubMed Abstract \| Publisher Full Text\n\nDaly KA, Lund EM, Harty KC, et al.: Factors associated with late smoking initiation in Minnesota women. Am J Public Health. 1993; 83(9): 1333–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDautzenberg B, de Souza Moura MA, Rieu N, et al.: [The e-cigarette disrupts other consumptions in Parisian teenagers (2012-2014)]. Rev Mal Respir. 2016; 33(3): 225–34. PubMed Abstract \| Publisher Full Text\n\nde Lacy E, Fletcher A, Hewitt G, et al.: Cross-sectional study examining the prevalence, correlates and sequencing of electronic cigarette and tobacco use among 11-16-year olds in schools in Wales. BMJ Open. 2017; 7(2): e012784. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nde Leon J, Diaz FJ, Rogers T, et al.: Initiation of daily smoking and nicotine dependence in schizophrenia and mood disorders. Schizophr Res. 2002; 56(1–2): 47–54. PubMed Abstract \| Publisher Full Text\n\nDoran N, Brikmanis K, Petersen A, et al.: Does e-cigarette use predict cigarette escalation? A longitudinal study of young adult non-daily smokers. Prev Med. 2017; 100: 279–284. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDutra LM, Glantz SA: Electronic cigarettes and conventional cigarette use among U.S. adolescents: a cross-sectional study. Erratum appears in JAMA 2014; 168 (7): 684. JAMA Pediatr. 2014; 168(7): 610–617. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDutra LM, Glantz SA: E-cigarettes and National Adolescent Cigarette Use: 2004-2014. Pediatrics. 2017; 139(2): pii: e20162450. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEastwood B, Dockrell MJ, Arnott D, et al.: Electronic cigarette use in young people in Great Britain 2013-2014. Public Health. 2015; 129(9): 1150–6. PubMed Abstract \| Publisher Full Text\n\nErnst M, Luckenbaugh DA, Moolchan ET, et al.: Decision-making and facial emotion recognition as predictors of substance-use initiation among adolescents. Addict Behav. 2010; 35(3): 286–9. PubMed Abstract \| Publisher Full Text\n\nEtter JF, Bullen C: A longitudinal study of electronic cigarette users. Addict Behav. 2014; 39(2): 491–4. PubMed Abstract \| Publisher Full Text\n\nFarsalinos KE, Poulas K, Voudris V, et al.: Electronic cigarette use in the European Union: analysis of a representative sample of 27 460 Europeans from 28 countries. Addiction. 2016; 111(11): 2032–2040. PubMed Abstract \| Publisher Full Text\n\nFarsalinos KE, Poulas K, Voudris V, et al.: Prevalence and correlates of current daily use of electronic cigarettes in the European Union: analysis of the 2014 Eurobarometer survey. Intern Emerg Med. 2017; 12(6): 757–763. PubMed Abstract \| Publisher Full Text\n\nFewell Z, Davey Smith G, Sterne JA: The impact of residual and unmeasured confounding in epidemiologic studies: a simulation study. Am J Epidemiol. 2007; 166(6): 646–55. PubMed Abstract \| Publisher Full Text\n\nGlasser A, Abudayyeh H, Cantrell J, et al.: Patterns of E-Cigarette Use Among Youth and Young Adults: Review of the Impact of E-Cigarettes on Cigarette Smoking. Nicotine Tob Res. 2018. PubMed Abstract \| Publisher Full Text\n\nGoniewicz ML, Gawron M, Nadolska J, et al.: Rise in electronic cigarette use among adolescents in Poland. J Adolesc Health. 2014; 55(5): 713–715. PubMed Abstract \| Publisher Full Text\n\nGreenland S: The effect of misclassification in the presence of covariates. Am J Epidemiol. 1980; 112(4): 564–569. PubMed Abstract \| Publisher Full Text\n\nGreenland S, Robins JM: Confounding and misclassification. Am J Epidemiol. 1985; 122(3): 495–506. PubMed Abstract \| Publisher Full Text\n\nGritz ER, Prokhorov AV, Hudmon KS, et al.: Cigarette smoking in a multiethnic population of youth: methods and baseline findings. Prev Med. 1998; 27(3): 365–84. PubMed Abstract \| Publisher Full Text\n\nHammond D, Reid JL, Cole AG, et al.: Electronic cigarette use and smoking initiation among youth: a longitudinal cohort study. CMAJ. 2017; 189(43): E1328–E1336. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHanewinkel R, Isensee B: Risk factors for e-cigarette, conventional cigarette, and dual use in German adolescents: a cohort study. Prev Med. 2015; 74: 59–62. PubMed Abstract \| Publisher Full Text\n\nHill A, Camacho OM: A system dynamics modelling approach to assess the impact of launching a new nicotine product on population health outcomes. Regul Toxicol Pharmacol. 2017; 86: 265–278. PubMed Abstract \| Publisher Full Text\n\nHornik RC, Gibson L, Lerman C: Prediction of cigarette use from six month prior electronic and combustible cigarette use for a U.S. national sample of 13-25 year olds [abstract POS5-30].Society for Research on Nicotine & Tobacco. 2016 Rapid Response Posters. 2016. Reference Source\n\nHuang M, Hollis J, Polen M, et al.: Stages of smoking acquisition versus susceptibility as predictors of smoking initiation in adolescents in primary care. Addict Behav. 2005; 30(6): 1183–94. PubMed Abstract \| Publisher Full Text\n\nHuh J, Leventhal AM: Progression of Poly-tobacco Product Use Patterns in Adolescents. Am J Prev Med. 2016; 51(4): 513–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJackson C: Cognitive susceptibility to smoking and initiation of smoking during childhood: a longitudinal study. Prev Med. 1998; 27(1): 129–34. PubMed Abstract \| Publisher Full Text\n\nJordan JN, McElroy JA, Everett KD: Smoking initiation, tobacco product use, and secondhand smoke exposure among general population and sexual minority youth, Missouri, 2011-2012. Prev Chronic Dis. 2014; 11: E113. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKalkhoran S, Glantz SA: E-cigarettes and smoking cessation in real-world and clinical settings: a systematic review and meta-analysis. Lancet Respir Med. 2016; 4(2): 116–28. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKaufman AR, Land S, Parascandola M, et al.: Tobacco use transitions in the United States: The National Longitudinal Study of Adolescent Health. Prev Med. 2015; 81: 251–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKozlowski LT, Warner KE: Adolescents and e-cigarettes: Objects of concern may appear larger than they are. Drug Alcohol Depend. 2017; 174: 209–214. PubMed Abstract \| Publisher Full Text\n\nLee PN: Summary of the epidemiological evidence relating snus to health. Regul Toxicol Pharmacol. 2011; 59(2): 197–214. PubMed Abstract \| Publisher Full Text\n\nLee PN: Epidemiological evidence relating snus to health--an updated review based on recent publications. Harm Reduct J. 2013; 10(1): 36. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLee S, Grana RA, Glantz SA: Electronic cigarette use among Korean adolescents: a cross-sectional study of market penetration, dual use, and relationship to quit attempts and former smoking J Adolesc Health. 2014; 54(6): 684–90. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLee P: Considerations Related to Vaping as a Possible Gateway into Cigarette Smoking. OSF. 2018. http://www.doi.org/10.17605/OSF.IO/Z3ST5\n\nLeventhal AM, Stone MD, Andrabi N, et al.: Association of e-Cigarette Vaping and Progression to Heavier Patterns of Cigarette Smoking. JAMA. 2016; 316(18): 1918–1920. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeventhal AM, Strong DR, Kirkpatrick MG, et al.: Association of Electronic Cigarette Use With Initiation of Combustible Tobacco Product Smoking in Early Adolescence. JAMA. 2015; 314(7): 700–707. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLevy DT, Borland R, Lindblom EN, et al.: Potential deaths averted in USA by replacing cigarettes with e-cigarettes. Tob Control. 2018; 27(1): 18–25. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLoukas A, Batanova M, Fernandez A, et al.: Changes in use of cigarettes and non-cigarette alternative products among college students. Addict Behav. 2015; 49: 46–51. PubMed Abstract \| Publisher Full Text\n\nLoukas A, Marti CN, Cooper M, et al.: Exclusive e-cigarette use predicts cigarette initiation among college students. Addict Behav. 2018; 76: 343–347. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLozano P, Barrientos-Gutierrez I, Arillo-Santillan E, et al.: A longitudinal study of electronic cigarette use and onset of conventional cigarette smoking and marijuana use among Mexican adolescents. Drug Alcohol Depend. 2017; 180: 427–430. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMajeed BA, Weaver SR, Gregory KR, et al.: Changing Perceptions of Harm of E-Cigarettes Among U.S. Adults, 2012-2015. Am J Prev Med. 2017; 52(3): 331–338. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcNeill A, Etter JF, Farsalinos K, et al.: A critique of a World Health Organization-commissioned report and associated paper on electronic cigarettes. Addiction. 2014; 109(12): 2128–34. PubMed Abstract \| Publisher Full Text\n\nMcRobbie H, Bullen C, Hartmann-Boyce J, et al.: Electronic cigarettes for smoking cessation and reduction. Cochrane Database Syst Rev. 2014; (12): CD010216. PubMed Abstract \| Publisher Full Text\n\nMiech R, Patrick ME, O'Malley PM, et al.: E-cigarette use as a predictor of cigarette smoking: results from a 1-year follow-up of a national sample of 12th grade students. Tob Control. 2017; 26(e2): e106–e111. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNational Academies of Sciences Engineering and Medicine: Public Health Consequences of E-Cigarettes. The National Academies Press, Washington DC. 2018. PubMed Abstract \| Publisher Full Text\n\nNutt DJ, Phillips LD, Balfour D, et al.: Estimating the harms of nicotine-containing products using the MCDA approach. Eur Addict Res. 2014; 20(5): 218–225. PubMed Abstract \| Publisher Full Text\n\nO'Connor RJ, Flaherty BP, Edwards BQ, et al.: Regular smokeless tobacco use is not a reliable predictor of smoking onset when psychosocial predictors are included in the model. Nicotine Tob Res. 2003; 5(4): 535–543. PubMed Abstract \| Publisher Full Text\n\nPearson JL, Johnson AL, Johnson SE, et al.: Adult interest in using a hypothetical modified risk tobacco product: findings from wave 1 of the Population Assessment of Tobacco and Health Study (2013-14). Addiction. 2018; 113(1): 113–124. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPhillips AN, Smith GD: Cigarette smoking as a potential cause of cervical cancer: has confounding been controlled? Int J Epidemiol. 1994; 23(1): 42–49. PubMed Abstract \| Publisher Full Text\n\nPicotte DM, Strong DR, Abrantes AM, et al.: Family and peer influences on tobacco use among adolescents with psychiatric disorders. J Nerv Ment Dis. 2006; 194(7): 518–23. PubMed Abstract \| Publisher Full Text\n\nPopova L, Owusu D, Weaver SR, et al.: Affect, risk perception, and the use of cigarettes and e-cigarettes: a population study of U.S. adults. BMC Public Health. 2018; 18(1): 395. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPrimack BA, Shensa A, Sidani JE, et al.: Initiation of cigarette smoking after e-cigarette use: a nationally representative study. Ann Behav Med. 2016; 50(Suppl 1): S68.\n\nPrimack BA, Soneji S, Stoolmiller M, et al.: Progression to Traditional Cigarette Smoking After Electronic Cigarette Use Among US Adolescents and Young Adults. JAMA Pediatr. 2015; 169(11): 1018–23. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nReed MB, Wang R, Shillington AM, et al.: The relationship between alcohol use and cigarette smoking in a sample of undergraduate college students. Addict Behav. 2007; 32(3): 449–64. PubMed Abstract \| Publisher Full Text\n\nRohde P, Lewinsohn PM, Brown RA, et al.: Psychiatric disorders, familial factors and cigarette smoking: I. Associations with smoking initiation. Nicotine Tob Res. 2003; 5(1): 85–98. PubMed Abstract \| Publisher Full Text\n\nSargent JD, Dalton MA, Beach M, et al.: Cigarette promotional items in public schools. Arch Pediatr Adolesc Med. 1997; 151(12): 1189–96. PubMed Abstract \| Publisher Full Text\n\nSavitz DA, Barón AE: Estimating and correcting for confounder misclassification. Erratum appears in American Journal of Epidemiology 1989; 130: 1260. Am J Epidemiol. 1989; 129(5): 1062–1071. PubMed Abstract \| Publisher Full Text\n\nScal P, Ireland M, Borowsky IW: Smoking among American adolescents: a risk and protective factor analysis. J Community Health. 2003; 28(2): 79–97. PubMed Abstract \| Publisher Full Text\n\nSelya AS, Rose JS, Dierker L, et al.: Evaluating the mutual pathways among electronic cigarette use, conventional smoking and nicotine dependence. Addiction. 2018; 113(2): 325–333. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSimons-Morton B, Crump AD, Haynie DL, et al.: Psychosocial, school, and parent factors associated with recent smoking among early-adolescent boys and girls. Prev Med. 1999; 28(2): 138–48. PubMed Abstract \| Publisher Full Text\n\nSoneji S, Barrington-Trimis JL, Wills TA, et al.: Association Between Initial Use of e-Cigarettes and Subsequent Cigarette Smoking Among Adolescents and Young Adults: A Systematic Review and Meta-analysis. JAMA Pediatr. 2017; 171(8): 788–797. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSoneji SS, Sung HY, Primack BA, et al.: Quantifying population-level health benefits and harms of e-cigarette use in the United States. PLoS One. 2018; 13(3): e0193328. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSpindle TR, Hiler MM, Cooke ME, et al.: Electronic cigarette use and uptake of cigarette smoking: A longitudinal examination of U.S. college students. Addict Behav. 2017; 67: 66–72. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSutfin EL, Reboussin BA, Debinski B, et al.: The Impact of Trying Electronic Cigarettes on Cigarette Smoking by College Students: A Prospective Analysis. Am J Public Health. 2015; 105(8): e83–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTemple JR, Shorey RC, Lu Y, et al.: E-cigarette use of young adults motivations and associations with combustible cigarette alcohol, marijuana, and other illicit drugs. Am J Addict. 2017; 26(4): 343–348. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTreur JL, Rozema AD, Mathijssen JJP, et al.: E-cigarette and waterpipe use in two adolescent cohorts: cross-sectional and longitudinal associations with conventional cigarette smoking. Eur J Epidemiol. 2018; 33(3): 323–334. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTzonou A, Kaldor J, Smith PG, et al.: Misclassification in case-control studies with two dichotomous risk factors. Rev Epidemiol Sante Publique. 1986; 34(1): 10–17. PubMed Abstract\n\nUnger JB, Soto DW, Leventhal A: E-cigarette use and subsequent cigarette and marijuana use among Hispanic young adults. Drug Alcohol Depend. 2016; 163: 261–4. PubMed Abstract \| Publisher Full Text\n\nUS Surgeon General: Preventing tobacco use among young people. A report of the Surgeon General. US Department of Health and Human Services; Public Health Service, Atlanta, Georgia. 1994. Reference Source\n\nWatkins SL, Glantz SA, Chaffee BW: Association of Noncigarette Tobacco Product Use With Future Cigarette Smoking Among Youth in the Population Assessment of Tobacco and Health (PATH) Study, 2013-2015. JAMA Pediatr. 2018; 172(2): 181–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWeiss JW, Garbanati JA, Tanjasiri SP, et al.: Effects of family functioning and self-image on adolescent smoking initiation among Asian-American subgroups. J Adolesc Health. 2006; 39(2): 221–8. PubMed Abstract \| Publisher Full Text\n\nWest R, Bauld L, O'Connor R, et al.: Expert reaction to meta-analysis looking at e-cigarette use and smoking cessation. Accessed: 2016 March. Reference Source\n\nWestling E, Rusby JC, Crowley R, et al.: Electronic Cigarette Use by Youth: Prevalence, Correlates, and Use Trajectories From Middle to High School. J Adolesc Health. 2017; 60(6): 660–666. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWills TA, Knight R, Sargent JD, et al.: Longitudinal study of e-cigarette use and onset of cigarette smoking among high school students in Hawaii. Tob Control. 2017; 26(1): 34–39. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nXu Y, Guo Y, Liu K, et al.: E-Cigarette Awareness, Use, and Harm Perception among Adults: A Meta-Analysis of Observational Studies. PLoS One. 2016; 11(11): e0165938. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhong J, Cao S, Gong W, et al.: Electronic Cigarettes Use and Intention to Cigarette Smoking among Never-Smoking Adolescents and Young Adults: A Meta-Analysis. Int J Environ Res Public Health. 2016; 13(5): pii: E465. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhu SH, Gamst A, Lee M, et al.: The use and perception of electronic cigarettes and snus among the U.S. population. PLoS One. 2013; 8(10): e79332. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "45926", "date": "28 Mar 2019", "name": "Lion Shahab", "expertise": [ "Reviewer Expertise Epidemiology", "Health Psychology", "Tobacco Control", "Psychopharmacology", "Neuroscience" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting and valuable contribution to the field on the hotly debated topic of whether or not e-cigarettes function as a gateway into subsequent combustible cigarette use. Of particular value here is the analysis of the impact of residual confounding on creating spurious associations in observational studies. However, there are a number of issues in the other analyses I would recommend addressing.\n\nAbstract: This is currently rather data free. I would suggest including some illustrative results here. Results: Particularly in relation to Tables 5-7, I wonder if it would be possible to condense results somewhat (or to move some of these results into supplementary files). It seems that it is not strictly necessary to include all modelled effects, especially since not all are discussed. This makes this section at times difficult to follow. A better way to structure the Tables might be to link these explicitly to the 5 conclusions presented on pages 9 and 10. Terminology: I am not sure that the term “gateway-out” is helpful. At the very least, it would not represent the only way to construe this concept insofar as it could be argued that in a hypothetical scenario the opposite to the “gateway-in” might be the avoidance of uptake of smoking in the first place (another “gateway-out” not related to smoking cessation where adolescents who would have smoked try an e-cigarette and never start smoking). It might therefore be helpful to clarify the difference between those gateway effects away from smoking Change in smoking trends: I am not entirely convinced by this analysis for a number of reasons. First, very little information is provided about the actual analysis undertaken. I would expect to see fitting parameters for various functions to model trends in smoking prevalence in the period prior to 2014. Are these assumed to be linear, logarithmic, polynomial etc.? More detail is required here. A more appropriate way to analyse the data would be to conduct segmented regression or timeseries analysis. Second, based on the current analysis, the problem (somewhat acknowledged on top of page 11) is that it might still be the case that e-cigarettes functioned as a “gateway in” since we do not know what smoking prevalence would have been in the absence of e-cigarettes (and may have been even lower). I don’t think this is likely but this issue should be acknowledged more explicitly, perhaps making it clear that such arguments would require “special pleading”, i.e. one would need to postulate additional environmental changes which could explain a continued decrease in smoking prevalence even in the presence of a gateway effect. Second paragraph on page 11 does not seem to add anything (it’s not clear how these cognitive measures in PATH relate to actual behaviour and so it’s quite weak evidence). Suggest dropping this. Table 9: The counterpart to A3 is not presented, i.e. dual use which results in reduction of combustible cigarettes. If there is a commensurate reduction in cigarette consumption associated with an increase in e-cigarette use (if dual users reduce cigarette consumption sufficiently) there may be beneficial health effects (so perhaps this effect should be added). In fact, perhaps a more nuanced discussion of dual use is needed on page 12, given that it may have both positive and negative effects and encompasses such a wide range of behavioural patterns. Please note that the statement on page 11 “The adverse effect for smokers switching to vaping instead of quitting smoking (A2) is not ideal but would still confer some benefits, given what is known regarding the reduced harm and toxicity of e-cigarettes\" is logically inconsistent. If a smoker intending to quit starts using e-cigarettes instead of quitting, there is no benefit (as e-cigarette use even if less harmful than smoking does not confer any benefit over stopping completely). The discussion on end of page 11 and top of page 12 feels rather qualitative. Either provide more quantitative analysis of modelled impact on health effects or move this to the discussion. You don’t really provide any primary data to support your conclusion ”Finally, introducing e-cigarettes should lead to a reduced risk of SRD.” This is rather speculative at this stage without more rigorous modelling as was done e.g. in the NASEM report.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4644", "date": "21 May 2019", "name": "Peter Lee", "role": "Author Response", "response": "Taking Dr Shahab’s numbered points: As the paper covers numerous issues and the conclusions are intended to be general, and as abstract length is limited, we left the abstract unaltered. We prefer not to change Tables 5 and 7. They are short and reducing them makes it harder to see effects of variation in the different parameters. We have abandoned the terms gateway-in and gateway-out. Gateway-in is now called the gateway effect, and gateway-out is explained more generally. A new additional file (4) gives fuller detail. The model used – linear in log (p/(1-p)) - was actually described originally, the additional information suggesting this is a reasonable approach. The discussion has been increased here. Rather than dropping some material in the section “Additional evidence on …”, we prefer adding a paragraph clarifying the data limitations. Table 9 and the related discussion have been amended along the lines suggested. We moved the last two paragraphs of results to the discussion." } ] }, { "id": "45925", "date": "12 Apr 2019", "name": "Sandra I. Sulsky", "expertise": [ "Reviewer Expertise Epidemiology", "population health modeling", "tobacco harm reduction." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis ambitious paper by Lee and colleagues seems to have several aims, and I fear that the paper suffers from including them all, rather than separating them out into two or more focused manuscripts.\nThe authors provide a review of the literature discussing what is currently known about the association between use of combusted and electronic tobacco products, they evaluate the likelihood of electronic products providing a gateway in to smoking based on seemingly informal, qualitative trend analyses, and they provide an important discussion about the potential for residual or otherwise uncontrolled confounding to explain or partially explain associations that have been noted in the literature. They also provide data from a series of simulations. The simulations demonstrate the interdependence among various tobacco use patterns in predicting population harm or benefit resulting from electronic tobacco products.\nThe most important points from the literature-based sections of the paper are a) the difference between predicted and observed smoking prevalence over time, which should reduce the fear that a large gateway to smoking effect is currently operating; b) the likelihood that misclassification and uncontrolled confounding play a role in the results reported by others, and probably lead to an overestimate of gateway-in effect.\nThe key findings from the simulations are a bit harder to uncover, due to a lack of detail in the methods section and some glossing over of assumptions. The authors found that a gateway-in effect is unlikely to be large enough to substantially affect population health. This is based on the assumptions that a) the current situation, in which smoking is far more prevalent than electronic cigarette use, will remain stable in at least the short run; b) the current higher likelihood of initiating tobacco use with cigarettes vs electronic products will remain stable in at least the short run; and c) the assumptions that the risks associated with electronic products are understood and are definitely lower than risks associated with smoking. If any of these input assumptions were varied, the simulation results would probably be different, and possibly much different. Some of the authors' strongest statements are due to differences between relative and absolute effects: Small percentages of a large underlying population (smokers) affect a large number of people, and large percentages of a small underlying population (electronic product users or initiators) affect a small number of people.\n\nApart from these conceptual concerns, it would strengthen the paper if the input into the simulations were better defined and justified.\nThe analyses presented in Table 8 seem to be the most informative of the simulations. Estimating the rate of the combination of smoking OR vaping OR dual use and comparing the prevalence in each year to the prevalence predicted by trends in smoking would give clues about the population level effects of e-products. Has overall use increased or stayed the same?\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4645", "date": "21 May 2019", "name": "Peter Lee", "role": "Author Response", "response": "Dr Sulsky makes three main points:The paper may be better as multiple focussed manuscripts. We did not attempt this, our main purpose being to cover a range of issues relevant to vaping. Note that two of us also recently published in F1000Research on “Investigating gateway effects using the PATH study”. Some key findings from the simulations are hard to uncover and some assumptions not clearly stated. We made numerous amendments to answer these points. It would also be useful to study trends in combined prevalence of vaping and/or smoking. As the paper’s interest concerns effects of vaping on smoking prevalence, and as interpretation of upward trends in combined use is problematic (they may well reduce risk) we did not change the paper here." } ] } ]	1	https://f1000research.com/articles/7-1915
https://f1000research.com/articles/8-1135/v1	19 Jul 19	{ "type": "Software Tool Article", "title": "Iron Hack - A symposium/hackathon focused on porphyrias, Friedreich’s ataxia, and other rare iron-related diseases", "authors": [ "Gloria C. Ferreira", "Jenna Oberstaller", "Renée Fonseca", "Thomas E. Keller", "Swamy Rakesh Adapa", "Justin Gibbons", "Chengqi Wang", "Xiaoming Liu", "Chang Li", "Minh Pham", "Guy W. Dayhoff II", "Linh M. Duong", "Luis Tañón Reyes", "Luciano Enrique Laratelli", "Douglas Franz", "Segun Fatumo", "ATM Golam Bari", "Audrey Freischel", "Lindsey Fiedler", "Omkar Dokur", "Krishna Sharma", "Deborah Cragun", "Ben Busby", "Rays H.Y. Jiang", "Gloria C. Ferreira", "Jenna Oberstaller", "Renée Fonseca", "Thomas E. Keller", "Swamy Rakesh Adapa", "Justin Gibbons", "Chengqi Wang", "Xiaoming Liu", "Chang Li", "Minh Pham", "Guy W. Dayhoff II", "Linh M. Duong", "Luis Tañón Reyes", "Luciano Enrique Laratelli", "Douglas Franz", "Segun Fatumo", "ATM Golam Bari", "Audrey Freischel", "Lindsey Fiedler", "Omkar Dokur", "Krishna Sharma", "Deborah Cragun" ], "abstract": "Background: Basic and clinical scientific research at the University of South Florida (USF) have intersected to support a multi-faceted approach around a common focus on rare iron-related diseases. We proposed a modified version of the National Center for Biotechnology Information’s (NCBI) Hackathon-model to take full advantage of local expertise in building “Iron Hack”, a rare disease-focused hackathon. As the collaborative, problem-solving nature of hackathons tends to attract participants of highly-diverse backgrounds, organizers facilitated a symposium on rare iron-related diseases, specifically porphyrias and Friedreich’s ataxia, pitched at general audiences. Methods: The hackathon was structured to begin each day with presentations by expert clinicians, genetic counselors, researchers focused on molecular and cellular biology, public health/global health, genetics/genomics, computational biology, bioinformatics, biomolecular science, bioengineering, and computer science, as well as guest speakers from the American Porphyria Foundation (APF) and Friedreich’s Ataxia Research Alliance (FARA) to inform participants as to the human impact of these diseases. Results: As a result of this hackathon, we developed resources that are relevant not only to these specific disease-models, but also to other rare diseases and general bioinformatics problems. Within two and a half days, “Iron Hack” participants successfully built collaborative projects to visualize data, build databases, improve rare disease diagnosis, and study rare-disease inheritance. Conclusions: The purpose of this manuscript is to demonstrate the utility of a hackathon model to generate prototypes of generalizable tools for a given disease and train clinicians and data scientists to interact more effectively.", "keywords": [ "Hackathon", "Data Science", "Ataxia", "Porphyria", "Rare Diseases", "Friedreich’s Ataxia", "Clinical Informatics", "Bioinformatics" ], "content": "Introduction\n\nHackathons are an effective avenue for the generation of software prototypes in the biomedical informatics space, several of which have been sponsored by the National Institutes of Health (NIH NCBI). A long-standing interest and active research programs on rare diseases, including Friedreich’s ataxia and porphyrias at the University of South Florida (USF), prompted us to modify the traditional NCBI Hackathon model and initiate a specific disease-focused hackathon1–3. Our event “Iron Hack” was named after the rare diseases upon which we focused.\n\nBecause of the diverse scientific and bioinformatic backgrounds of hackathon participants, the organizers felt it necessary to have a symposium on rare iron-related diseases, and specifically ataxia and porphyrias, in the early part of each of the three days. During the symposium, renowned scientists, clinicians and investigators in rare iron-related disease research covered the major aspects of Friedreich’s ataxia, porphyrias and sideroblastic anemia and emphasized pressing questions that need to be addressed for advancement of the field. As a result of the hackathon, we developed resources that are relevant not only to rare iron-related diseases, but also to other rare diseases and some general bioinformatics problems. The objective of this report is to demonstrate the utility of a hackathon model to develop generalizable tools for evaluation, diagnosis, and management of a given disease.\n\nRare diseases have a large impact on the population, with 7,000 orphan diseases collectively affecting about 1 in 10 to 20 people4. These diseases place a heavy burden on patients and families, with a diagnosis taking up to ten years to identify, if ever. Limited patient-numbers and resources for each disease severely hamper research, prognoses, diagnoses, and treatments. Although quite disparate in symptoms, these conditions all stem from problems with iron metabolism or heme synthesis. A brief description of these rare diseases is provided below to provide context for the “Iron Hack” team projects.\n\nBiochemical basis and clinical manifestation. Porphyrias are a group of rare metabolic disorders caused by malfunction of the enzymes involved in heme biosynthesis2,3,5. There are eight different types of porphyria, each of which arises from mutation(s) in the genes for each of the eight enzymes of the heme biosynthetic pathway2,3,6,7. With the exception of porphyria cutanea tarda (PCT), all other seven porphyrias are inherited as autosomal-dominant, autosomal-recessive, or X-linked traits2,6,8,9.\n\nAll porphyrias are characterized by the accumulation and excretion of porphyrins or porphyrin precursors, though each type has disparate clinical manifestations (including neurovisceral and/or cutaneous symptoms) depending upon which enzyme of the pathway is defective10. In general, neurological disturbances are manifested in the form of acute attacks (e.g., extreme abdominal and chest pain, vomiting, confusion, constipation, fever, high blood pressure, low blood sodium levels, and seizures), while photosensitivity is at the root of cutaneous manifestations (e.g., skin blistering, redness, scarring, and pain when exposed to the sun)3,9,11,12. Suggested treatments and disease-outcomes also vary with the particular pathway-defect10,13–19.\n\nDiagnosis. Acute porphyrias can prevail undiagnosed for 10–15 years following the onset of symptoms19. Perhaps not surprisingly, diagnosis of porphyrias remains challenging: they are rare, and their symptoms are nonspecific, often mimicking other, more common disorders11. Thus once porphyria clinical symptoms are recognized, biochemical laboratory testing should be performed to identify the specific type of porphyria10. Genetic testing, normally targeted gene sequencing, becomes critical to define the mutation(s) in a family and make genetic counseling possible. However, diagnosis of porphyrias are often overlooked due to, a large extent, the difficulty in performing the specific biochemical assays and absence of specialized porphyria centers, particularly in developing countries20–22.\n\nPrevalence. In the United States, porphyrias collectively afflict fewer than 200,000 people, with similar prevalence in the European Union4,10,23. Estimates of porphyria prevalence vary by type, with values of 1 in 10,000 for the most common type of porphyria, PCT (OMIM 176090) and 1 in 1,000,000 for congenital erythropoietic porphyria (CEP; OMIM 263700)10. Many people with a genetic mutation associated with the disease never experience signs or symptoms, a phenomenon known as incomplete or reduced penetrance24.\n\nBiochemical basis and clinical manifestation. Friedreich’s ataxia (FRDA, FA; OMIM 229300) is a rare autosomal-recessive disease associated with progressive spinocerebellar ataxia, cardiomyopathy, scoliosis, diabetes, and vision and hearing impairment1,25. Most FRDA patients are largely asymptomatic during the first 5 – 10 years of life. But, with advancing gait and limb ataxia, they require use of a wheelchair and are unable to perform daily activities independently, often during adolescence25,26.\n\nSymptoms result from reduced synthesis of the mitochondrial protein frataxin, an iron chaperone that, by shielding this metal, prevents the production of reactive oxygen species (ROS) and renders it bioavailable as ferrous iron26–29. When frataxin levels are low, iron accumulates in the mitochondria, largely in an oxidized and insoluble form30,31. The accumulated iron can participate in Fenton chemistry leading to formation of extensive reactive oxygen species (ROS) that cause damage and cell death. While there is a general agreement that frataxin is critical for mitochondrial iron metabolism and cellular iron homeostasis, its precise biological role remains a controversial matter26,32,33. Involvement of frataxin in 1) iron delivery to the iron–sulfur cluster assembly and repair machinery, 2) repair of oxidatively inactivated [3Fe–4S] aconitase to yield an active enzyme, 3) delivery of ferrous iron to ferrochelatase for heme biosynthesis, 4) detoxification of iron by catalyzing the oxidation of Fe(II) to Fe(III) and storing the metal as a ferrihydrite mineral within structurally organized frataxin oligomers are among the reported functions ascribed to frataxin26,31,34–41. Despite the lack of consensus, frataxin and heme biosynthesis are linked. Frataxin may participate in the assembly of the [2Fe-2S] cluster, an essential cofactor for an active human ferrochelatase, the terminal enzyme of the heme biosynthetic pathway42–44. Alternatively, by maintaining and chaperoning iron in a reduced form, frataxin may donate Fe(II) to ferrochelatase, which has strict physiological specificity for Fe(II) as substrate37,39,44–46. Clearly, a combination of these two functional possibilities cannot be ruled out.\n\nPrognoses and treatment. Presently, there is neither a cure nor a U.S. Food and Drug Administration (FDA)-approved treatment for FRDA47. Advances in understanding the underlying mechanism of FRDA, in particular the recognition that frataxin deficiency is the root cause of FRDA, have prompted the development of therapeutic strategies. Since increased oxidative stress and mitochondrial respiratory chain dysfunction have been associated with the pathogenesis of FRDA, antioxidants and inhibitors of free radical formation (e.g., idebenone, L-acetylcarnitine, resveratrol, and RT001 and other deuterated polyunsaturated fatty acids) have been assessed as a promising treatment option47–50. Iron chelators, such as deferiprone, have been considered as a therapeutic approach of FRDA by controlling iron accumulation and decreased frataxin synthesis29,51. Regulation of frataxin gene expression by increasing either histone acetylation or transcription of the frataxin gene, represents yet another treatment possibility being explored52,53. Histone deacetylase inhibitors reverse or, at least, diminish silencing and the reduced transcription of the frataxin gene observed in FRDA patients54. While protein interferon- increases transcription of the frataxin gene and consequent production of the frataxin, its therapeutic benefit remains to be established52,55. Frataxin gene replacement is also being developed for as a potential treatment for FRDA. Because of the scope of this report, an evaluation of the therapies for FRDA can be neither extensive nor complete47.\n\nPrevalence. FRDA affects about 1 in 150,000 individuals of Caucasian descent and accounts for 50% of overall cases of hereditary ataxia and for 75% of those with onset before age 251,25,56,57.\n\nThe robust evidence suggesting these devastating, rare diseases named porphyrias can largely be pinpointed to dysfunction in a single pathway of eight enzymes, caused by mutation(s) inherited in well-understood, classical Mendelian patterns made them an attractive case for potentially-impactful tool-development. However, even in these seemingly straightforward cases of diseases exhibiting classical Mendelian inheritance, disease phenotypes are not entirely explained by the presence of known pathogenic variants. The discrepancy between the low penetrance of symptomatic patients for autosomal-dominant acute intermittent porphyria (AIP; OMIM 176000) and the high frequency of pathogenic mutations led Chen et al. to propose that predisposing- or protective-modifier genes alter expression of the AIP phenotype58. Indeed, a small number of modifier-genes, regulatory and pathophysiological mechanisms have since been identified to contribute to onset of porphyrias, though these findings remain insufficient to explain the disease-penetrance puzzle8,20.\n\nIn rare diseases resulting from trinucleotide copy-number repeat-variation, such as FRDA, there is some degree of relationship between severity of disease phenotype and copy-number57,59. In FRDA, expanded trinucleotide (GAA) tracts in intron 1 of the FXN gene, commonly between 600 and 900 repeats, result in pathologically decreased levels of frataxin60–62. However, number of trinucleotide repeats are not reliably predictive of disease severity, further suggesting the importance of as-yet unknown modifying genes or environmental factors that may contribute to disease outcomes47.\n\nThe number of disease-phenotypes entirely decided by single-gene variants are in the minority63. Most inherited diseases are likely to have a more complicated etiology determined by some combination of genomic variants, impacted by myriad environmental factors as well.\n\nWe organized Iron Hack to address these challenges, including the great need for genomics tools to handle rare-disease data, such that new data-mining concepts and computational tools could be developed and further adapted to serve the rare-disease communities. We established five Iron Hack teams to develop five computational-tool prototypes broadly focused on (1) exploration of consumer-genomics data, (2) large-scale RNAseq data mining, (3) genomic data visualization, (4) rare-disease variants discovery, and (5) genotype-to-phenotype mapping. These team-efforts have led to the convergence of iron-research communities and genomics data-science researchers to produce promising computational tools, strengthened through an iterative process of soliciting ideas and feedback from domain experts.\n\nThe remainder of this report is organized into subsections by project, beginning with a detailed description for the five projects, the motivations behind them and the gaps they seek to fill. We next describe the methodologies and implementations of the projects into usable software applications, how to operate the software applications, and results produced using the software applications. Finally, we discuss the pros and cons of this new highly-interdisciplinary and community-driven twist on more traditional hackathons.\n\n\nProject descriptions and goals\n\nUniting People Working Against Rare Diseases (UPWARD) will be a Health Insurance Portability and Accountability Act (HIPAA)-compliant database which will allow people with rare diseases to declare interest in participating in research studies, and subsequently share their personal disease stories, clinical symptoms, and consumer genetic testing data with researchers and clinicians. Figure 1 shows that, as consumer-genetic testing data are submitted, they are analyzed alongside 43 porphyria-related pathogenic SNPs currently held in UPWARD. A set of statistical computation and machine learning methods can be used downstream to parse out the novel modifiers of diseases as well as the interactions of genetic loci underlying pathologies. This information with be compiled and analyzed within UPWARD using, in part, a program which identifies all rare disease-related pathogenic or likely-pathogenic Single Nucleotide Polymorphisms (SNPs) that are currently included on SNP microarray chips used by common consumer genetic testing companies. Table 1 shows that variants sourced from ClinVar, a crowdsourced genotype-phenotype database hosted by NCBI, against consumer-genetics data sourced from Illumina OmniExpress and GSA microarray chips used by Ancestry and 23andMe. These resulting 43 variants will be used in analysis of patient-submitted consumer-genomics results. The goal of this platform is to facilitate data-driven discovery of rare-disease determinants, such as modifiers that affect penetrance, by leveraging the growing data of consumer genomics. To facilitate use of this database, UPWARD has focused its tools to benefit people living with porphyria, and porphyria research as a whole.\n\nUPWARD opens with a web interface designed to clearly communicate research and advocacy goals to the public, request consent and gather data in a HIPPA-compliant manner.\n\nUPWARD includes a tool built to map highly-pathogenic and likely-pathogenic porphyria-associated variants.\n\nWhen people with porphyria access UPWARD, they are met with a survey built to collect consent, contact information and disease-associated information, such as clinical symptoms, genetic and environmental data, including targeted questions concerning environmental factors suspected to trigger acute porphyria attacks. Participants are given the option to share this survey with family members and friends, both those with and without porphyria symptoms. Although family members and friends without symptoms at the time of the survey will likely never develop symptoms (due to the low penetrance of porphyria-associated mutations), we seek to identify modifying genes and environmental factors that contribute to the phenotype through comparing genotypes of these individuals with those of people reporting latent and active porphyria64. We plan to explore the possibility of recruiting participants by adding UPWARD links to the SNPedia research database, as well as through collaborating with porphyria advocacy and patient-education groups, and clinical partners.\n\nMany different mutations can contribute to the onset and progression of porphyria65. We designed a method to search for underlying genetic variants associated with symptoms of congenital erythropoietic porphyria (CEP). The first diagnostic steps to confirm CEP often happen after referral to a genetic counsellor, who recommends targeted screening for a panel of known-pathogenic porphyria-associated SNPs. In cases where no known-pathogenic variants are found, whole-exome sequencing may be recommended of both the patient and their parents to catalog variation in the symptomatic person versus their asymptomatic parents. These variants filtered from the parent-child “trio” data can then be annotated with available disease-associated information, if any, using existing tools (such as dbNSFP and WGSA)66,67. With our Variants Discovery tool, we aimed to generate a workflow which operates on trio-data to identify, categorize and then rigorously assess candidate disease-causing mutations in cases where the underlying mutation is unknown, modeled after existing workflows for whole-exome sequence analysis (Figure 2)68.\n\nAbbreviations: dbNSFP, database for nonsynonymous SNPs’ functional predictions; WGSA, whole genome sequencing annotator; HGMD, Human gene mutation database; eQTL, expression quantitative trait loci.\n\nTier 1 variants are known disease-causing mutations in known disease-causing genes. Tier 2 variants are uncharacterized de novo mutations predicted to be damaging (see Methods) in known disease-causing genes. Tier 3 variants are uncharacterized, damaging, inherited mutations in known disease-causing genes (parents are not affected). Tier 4 variants are functional mutations with unknown significance in known disease-causing genes. Tier 5 variants are damaging mutations in the extended gene list (e.g. those genes associated with symptoms of disease). Candidate disease-causing variants are categorized into five evidence-based tiers, where Tier 1 variants are known-pathogenic and have the highest support. We intend to expand this workflow so that it might be used to assist in the diagnosis of patients with other difficult-to-identify conditions.\n\nThe fields of biology and medicine have undergone swift changes to the manner in which ribonucleic acid (RNA) can be studied using deep-sequencing techniques to investigate expression-differences in possible RNA species that may be associated with deleterious disease outcomes69. RNA-Seq technology has revolutionized detection and analysis of aberrant RNA transcripts associated with disease69.\n\nIn rare disease research in particular, obtaining sample-sizes enabling confident identification of disease-associated transcripts is a considerable challenge. The amount of RNA-seq data contributed to NCBI’s Gene Expression Omnibus (GEO), a public repository for functional genomics data, is increasing at a rapid pace. A simple query for “Expression profiling by high throughput sequencing” yielded 14,200 unique datasets as of March 6, /2019. The availability of these massive quantities of data creates an open opportunity in many research areas for meta-analyses using these published datasets to strengthen analytical power. Our massive parallel-sequencing analysis tool, MassiveSeq, provides an opportunity for researchers and bioinformaticians to easily extract and process meaningful information (such as quantitative gene expression, novel transcripts and their isoforms, alternative splice-site variants, SNPs or copy-number variation) from these large datasets to evaluate the associations between biological processes, gene expression and disease outcomes. MassiveSeq automates downloading and processing of large-scale RNA-seq datasets with the aim of easing computational time and complexity70–73.\n\nMassiveSeq differs from conventional, comprehensive RNA-seq pipelines in that it combines multiple RNA-seq datasets to increase analytical power (Figure 3)74–76. The search is confined to samples meeting the criteria, e.g., disease, library source (genomic, transcriptomic or metagenomic), platform (Illumina, PacBio), or instrument (Genome Analyzer, Hiseq, Nextseq). MassiveSeq additionally allows exploration of novel clustering methods to enable meta-analysis of differential gene expression. Initial steps in processing raw sequencing reads on even a single, traditional dataset are often computationally intensive, and obtaining additional publicly-available RNAseq datasets at a massive scale for such processing is resource-consuming as well. MassiveSeq takes raw-sequencing data (fastq format) automatically streamed from NCBI’s Short-Read Archive (SRA) as input, using a GEO query specifying parameters such as disease and experimental type (e.g., high-throughput RNAseq). Datasets can be further filtered as needed. The MassiveSeq pipeline next utilizes dockerized HISAT2 (version 2.1.0) and StringTie (v1.3.5) to enable automated, parallel processing of each experiment77,78. Reads are automatically streamed directly from SRA, mapped to a reference genome, assembled into transcripts--including novel splice-variants--and annotated in parallel within each dataset. The MassiveSeq pipeline allows uniform processing of multiple, independent RNAseq datasets, enabling powerful identification of differentially expressed genes and transcripts associated with diseases of interest. We applied MassiveSeq to 99 Friedreich’s Ataxia SRA datasets to identify disease-associated transcripts for Iron Hack70–73.\n\nThe pipeline takes metadata from the Sequence Read Archive (SRA) and parses it for quality control (QC). The primary work takes place in a custom snakemake script that aligns sequences with Hisat2 and then quantifies transcripts with Stringtie in a parallelized fashion across available machines and cores.\n\nAbnormal gene-expression patterns can cause a broad range of diseases. However identifying abnormally-expressed genes and correctly interpreting expression data across experiments can be complicated by inconsistencies in gene-expression normalization strategies, as well as inadequate filtering of noisy data. Here, we developed an algorithm to rapidly identify genes with abnormal gene expression patterns in samples of interest (e.g., disease-presenting patient) as compared to controls (Figure 4). This method was built utilizing ~2000 RNA-Seq datasets publicly available on GTExPortal79. The package implements three commonly used RNA-Seq normalization methods: Fragments per kilobase of transcript per million mapped reads (FPKM), transcripts per million mapped reads (TPM) and differential gene expression analysis based on the negative binomial distribution (DEseq). A Gaussian-mixture model is utilized here to remove RNA-Seq noise and the DE-Seq method is finally implemented to capture abnormally expressed genes corresponding to query tissue. Simulation data were generated to test algorithm performance, and we intend to expand this system so that it might be used to assist in the diagnosis of patients with difficult to identify conditions80.\n\nAfter RNA-Seq is performed on a patient sample, the program searches the Genotype-Tissue Expression Project (GTEx) database for RNA-Seq data from the specific tissue potentially associated with the disease. Three methods are used for RNA-Seq normalization (Fragments per kilobase of transcript per million mapped reads (FPKM), transcripts per million mapped reads (TPM) and Differential gene expression analysis based on the negative binomial distribution (as implemented in DESeq)), and the data were fit to a Gaussian mixture model to remove noise within samples. The differentially expressed genes in the patient sample are finally captured by using the R program DESeq.\n\nDisease-phenotypes are unlikely to be entirely explained by the presence of single pathogenic variants. Pleiotropy, modulating genes and combinatorial effects are the rule, rather than the exception; however assessing combinatorial effects underlying disease quickly becomes computationally expensive, with a practically-infinite number of variant-combinations that could be assessed. We developed a tool-set to enable thoughtful reduction of variants to feasibly assess the role of modifying genes in rare diseases such as Friedreich’s ataxia.\n\nMost of the alleles a person inherits are unlikely to be involved in modulating the disease phenotype, and models incorporating many extraneous variables are unnecessarily cumbersome and perform more poorly than models incorporating domain-specific feature-selection. Therefore the first step of our pipeline was to reduce the disease-associated variant search-space to genes fitting a profile of interest.\n\nWe focused on broadly-applicable features of likely disease-causing variants (as opposed to disease-specific features) for our first layer of feature-selection in this first iteration of our pipeline. Input variant-call data are filtered based on the likelihood that any particular variant is deleterious (as predicted by Polyphen-2 scores) and by residual variation inheritance scores81,82. As features-of-interest for disease-associated variants should change depending upon the particular disease, phenotype, and available domain-specific knowledge, the feature-selection component of our tool is intended to be easily extendable for investigating the combinatorial contributions of multiple variants to disease phenotypes by any number of characteristics. We incorporated two existing annotation packages (Open-CRAVAT (version 1.4.0) and ANNOVAR version (On 2018Apr16)) to thoroughly annotate available information for each variant, any of which can be filtered on in the feature-selection module83. Highly ranked variants are then assessed for their contribution to disease-phenotype via the equally modular “analysis” part of our pipeline. Our current analysis module utilizes the APRIORI algorithm to detect variant co-occurrence relationships with disease, though the output from the feature-selection module is in a common format to facilitate application of other machine-learning approaches to identifying combinatorial interactions, all implemented via a simple web-user interface84.\n\nWe developed this pipeline with modularity being a primary goal. The APRIORI algorithm is currently implemented to identify genes that frequently co-occur in the feature-selected set of genes. Future work will implement tools that check for over-representation of gene ontology terms among the genes determined to have deleterious alleles.\n\n\nMethods and implementation\n\nKey concepts informing methods and implementations of each project are described below.\n\nTo build a database of pathogenic or likely pathogenic SNPs, we sourced Rapid Stain Identification Series (RSID) information from the Illumina OmniExpress & Illumina Global Screening Array (GSA) microarray chips (used by Ancestry and 23andMe respectively), then filtered out non disease-associated genes using the NCBI ClinVar database. For specific application to porphyrias, we selected all genetic polymorphisms annotated to be associated with any of the porphyrias, as well as their associated RSID, SNP location in the genome, and degree of pathogenicity85,86. Participants’ raw genomics data and environmental data are stored in a non-relational database, which has been proven to be more efficient than relational databases for storing and accessing genomic data87,88. A secure, HIPAA-compliant human subject meta-information database will be built as part of the next iteration of development89,90. A secure, HIPAA-compliant human subject meta-information database will be built as part of the next iteration of development89,90. At that time, the database will be expanded to capture the following information: 1) patient-reported phenotype and symptom information of people identified as potentially carrying a pathogenic or likely pathogenic variant in a porphyria gene and 2) people with a clinical diagnosis of porphyria, as well as de-identified information on their family members to try to capture data on asymptomatic people.\n\nOur system currently consists of a cloud-database built on MongoDB Community Edition, and a web server run through NGINX to accept input data from participants. The entire system is containerized and orchestrated by Docker Compose for ease of replication and to enable application to other diseases.\n\nOur pipeline categorizes patient variant-data into five tiers of pathogenic certainty based on quality of evidence, the logic of which is broadly outlined in Figure 2. The pipeline accepts dbNSFP or WGSA-annotated patient variant-files (in tab-delimited format, one variant per line). Annotated variants are first checked against known disease-associated variant databases, namely HGMD and ClinVar, to identify any previously reported pathogenic mutations matching the patient phenotype; these known, disease-causing variants in known disease-causing genes are categorized into the most confident classification, Tier 1. All variants not represented in the HGMD and ClinVar databases are next checked to see if they are located in genes that are involved in known disease-associated pathways. Variants in disease-associated pathways are then evaluated for probability of being deleterious (with start-loss, stop-gain, essential splicing variant, frameshift, indel or missense-mutations being highly likely to be deleterious). Damaging variants in known disease-causing genes are then contrasted against variant files from the non-affected parents to distinguish de novo (Tier 2, uncharacterized, damaging de novo mutation in known disease-causing gene) from inherited (Tier 3, uncharacterized, damaging inherited mutation in known disease-causing gene) variants. Non-deleterious variants are not considered further.\n\nDamaging variants not occurring in known disease-causing genes themselves, but mapping to known expression-Quantitative Trait Loci (eQTLs, loci associated with expression-changes in transcripts from known disease-causing genes), are stratified into Tier 4. We report all other damaging variants in or associated with genes that are related to symptoms of the disease as Tier 5 (damaging mutation in the extended gene list). Any other known disease-causing mutations associated with unrelated diseases are additionally reported in an extended report to allow for possible incidental or secondary findings.\n\nWe planned the main snakemake (version 5.4.2) to automate dispatching of jobs depending on the available cores and memory of a machine91. Here, the core steps involved Hisat2 for alignment, followed by Stringtie (v1.3.5) for transcript annotation and de novo annotation. Finally, reads were quantified by using featureCounts to measure at the exon level from the subRead (version 1.6.3) package92. This quantification pipeline follows a common, recently published protocol on Stringtie and Hisat293. It allows for both known as well as novel isoform transcripts to be identified and measured.\n\nOnce the gene counts were fully quantified for each sample, we analyzed the overall dataset comprised of all 4 studies using the R package metaSeq (version 1.22.1)94. This package adapts the non-parametric NOISeq method for differential RNA-seq analysis to allow for multiple studies in a meta-analysis framework95.\n\nWe used the gene counts with GSVA (Gene Set Variation Analysis, version 1.30) to estimate per-same GSEA (Gene Set Enrichment Analysis) pathway enrichments for the 50 hallmark datasets from MSigDB96,97. We used these pathway enrichments as features (a binary up- or down-regulated pathway) for a deep learning model, along with the remaining gene estimates. We used the fast.ai library to construct a Convolutional Neural Network (CNN)98. One of the novel features of fast.ai, especially for our data, is that it facilitates rapid construction of neural networks with tabular data via embeddings similar to Word2Vec99. The training split was 70/30; afterwards the CNN was trained for 5 epochs (cycles of the data), with a learning rate of 0.1.\n\nWe are also in the early stages of adapting the GATK RNA-seq best practices to this pipeline so that we can rapidly call variants on these samples100. Our workflow for the procedures, and methods used can be found in Figure 3.\n\nThis package is designed to detect abnormal genes exhibited differential expression compared with normal tissue cell. For each patient RNA-Seq result, we first download the gene expression level from normal tissue same as patient tissue. Then for multiple gene expression RNA-Seq samples available in website GTExportal (https://gtexportal.org/home/), three normalized methods including TPM, FPKM and DEseq are available. The Gaussian-mixture model is utilized to remove RNA-Seq noise. The basic idea is to use the EM algorithm to find two best fitted Gaussian distribution and only maintain the distribution with relative higher mean value as a true signal. After noise reduction, DEseq algorithm is used to identify significant up or down regulated genes in patients.\n\nThis tool is envisioned to be most useful for analyzing variant-call data from networks of families affected by rare disease (Figure 5). Currently the input data are variant-call files (.vcf) obtained from patients with a rare disease; future versions will incorporate additional genomic information from family members. The user uploads patient .vcf files through a web-app interface, then selects features on which to filter (currently a maximum and minimum score for residual variance-intolerance, a measure of gene-tolerance to variation based on population allele-frequencies). The web interface can be run locally to keep patient data secure. The back-end next runs the files through CRAVAT and ANNOVAR to assemble annotation information on all variants compiled from multiple databases that the user selects upon install (including ClinVar, Pubmed, etc). These annotation data are then filtered according to user-specification using the feature selection module.\n\nInput data are variant-call files in .vcf format collected from patient samples. The feature-selection module collects all available annotation information for each identified variant, then narrows down to variants most likely to be associated with the phenotype based on user-specified parameters. These feature-selected variants are then analyzed for combinatorial contribution to the disease using the tools in the analysis module. The output of the analysis modules are tables and graphs that summarize the results.\n\n\nOperation\n\nOperation can be performed on a computational cluster with multiple cores. The system can use a Lustre parallel file system for fast Input and Output. Remote mounting onto the cluster should be available for flexible data access and movement.\n\nThe only requirement to build this system is having Docker and Docker Compose installed on your machine. For instructions on running the system refer to the associated GitHub readme at bit.ly/UPWARD19101,102.\n\nGitHub readme and description available at https://bit.ly/2FGqkv7103,104.\n\nFor full instructions on how to clone and implement the code, please refer to:\n\nThe MassiveSeq github repository: https://bit.ly/2HKA61y105,106.\n\nRunning the web app requires the installation of Ruby, R, and Python on the server. The instructions for installing Ruby on Rails on Windows 10, Ubuntu, and OS X can be seen here, and should be similar for different OS versions107. Install Python 3 from here and install R from here108,109. The web application is available on GitHub at https://bit.ly/2V3Hpo2 and instructions for installation are detailed on the ReadMe110,111.\n\nGitHub readme and description available at https://github.com/NCBI-Hackathons/pheno_geno_ataxia112,113.\n\n\nLessons learned\n\nThroughout this process we identified several areas where improvements could be made for future disease-focused hackathons. A few of these are described below.\n\n1) We were successful in prototyping for a specific disease.\n\n2) It was helpful to learn more about the diseases and current problems that need to be solved before starting the projects.\n\n3) If trying to solve a clinical problem, such as how to improve and speed up the rate at which patients receive a diagnosis for rare diseases, include clinicians as part of the group.\n\n4) It may also be advantageous to have the didactic presentations about the diseases in advance of the hackathon such that everyone has a basic understanding of the issues and disease symptoms and time for brainstorming.\n\n5) Having a team meeting prior to the hackathon to assign roles and discuss overall flow for each day was helpful.\n\n6) Providing literature to read about the disease/genetic condition was also useful background for preparing for the hackathon.\n\n7) Having two leads on each team increased efficiency, as each could take turns fielding questions from less-experienced team-members while the other could keep the hacking on-task for the day.\n\n8) A few things some specific teams learned\n\na) UPWARD: Forming a team composed of people with a variety of training backgrounds (e.g. clinicians, researchers, organizers, computer scientists, biologists, geneticists, etc.) brings strength and utility to team ideas and project results. Additionally devoting a portion of the first day or meeting period to brainstorming, idea proposal, and arguments facilitates the formation of a plan which team members are able to agree upon and work towards while reducing the chance of a schism further down the road.\n\nb) MassiveSeq: We initially overscoped/planned on a completely different toolset. Scaling back to a core set of tools that we were comfortable with made completing the project feasible. People were coming in from really different backgrounds (choice of programming language, familiarity with genomics data formats, etc.) and in retrospect we would have liked to have planned a bit more for some specific tasks.\n\n\nResults\n\nAfter identifying porphyria-related pathogenic SNPs currently included on Illumina OmniExpress and Illumina GSA microarray chips, 43 porphyria-related pathogenic SNPs were found. This list (presented in Table 1) will be maintained and updated at the UPWARD GitHub repository (labeled as pathogenic SNPs.csv)101. This list will be compared against participant-submitted consumer genomics test results within UPWARD once the project is reviewed and approved by the Institutional Review Board (IRB).\n\nWe examined the genetic basis with the following examples of porphyria genetics. Using our developed pipeline, we successfully identified several candidate SNPs (in Tier 1, Tier 2 and Tier 4) that were previously unnoticed in a porphyria patient in a clinical setting. These SNPs are located in genes known to cause different kinds of porphyria, e.g. UROS and CPOX genes. The discovered SNPs can, based on the prediction using our pipeline, affect the transcription of the candidate genes, their translation or both. All of these possibilities would result in functional abnormalities of the final gene product. For example, two of the identified SNPs were eQTL (expression quantitative trait loci), which led to significantly decreased expression level of the UROS gene (Figure 6 and Figure 7).\n\nThere is significant down-regulation of UROS gene associated with this variant in all tissues (except ovary). NES: normalized effect size.\n\nThere is significant down-regulation of UROS gene associated with this variant in all tissues (except ovary). NES: normalized effect size.\n\nThese variants examples demonstrated that our pipeline can help physicians and/or clinical geneticists quickly filter out the vast majority of neutral variants and report the remaining variants in clinically meaningful tiers to facilitate further experimental validation and explanation.\n\nUsing Metaseq we identified over 2000 genes upregulated in Friedreich’s Ataxia patients compared to controls (Figure 8). However, we emphasize that this analysis was merely a proof of concept, and further work needs to be done to explore methods and techniques for standardizing phenotypes (data harmonization) alongside the meta-analysis itself in Friedreich’s Ataxia. (Figure 8).\n\nDistribution of significance in downregulated genes from metaseq analysis; no genes were significant at 0.05 threshold.\n\nWe used fast.ai to train a CNN on an embedded feature-space of these counts as well as 50 gene-set enrichment features from Msigdb (see Methods). The trained model had an overall accuracy of 0.75, which seems promising given the number of features and low number of samples for training.\n\nMassiveSeq’s implementation of HISAT2 and StringTie identified novel-isoform transcripts in various samples. We focused our analyses on the FXN gene, as trinucleotide GAA-repeats at this locus are causative of FRDA. We identified multiple novel-isoform transcripts within 1kb up and downstream of FXN in affected, unaffected and carrier-patients (Table 2). We were able to visualize the truncation of the FXN transcripts from the above samples using IGV. Shallow read-coverage of the whole transcriptome from this particular study made it difficult to confirm the reliability of the identified transcript truncation.\n\nFor each input patient RNA-Seq data, the RNA-Seq data related to query tissue are extracted from the database. The available tissues and number of RNA-seq data are listed in Table 3.\n\nFor convenience, brain tissue is selected here for the following discussion. Considering different sequencing depth for each sample, we provide three methods for data normalization: DEseq, FPKM, and TPM. As there are a large number of samples, we used uniform sampling to select n genes for visualization. As shown in Figure 9, n = 20 genes are shown here to compare different normalization methods. DEseq normalization results show relative lower fluctuation compared with the other two methods (F-test p < 2.2e16), indicating better performance of DEseq. Except for Bladder, Cervix, and Fallopian, most tissues in our database exhibit large RNA-Seq sample number. Therefore, a method is required to select the data with relatively high signal/noise ratio. A Gaussian-mixture model is fit for each gene and returns the posterior probability to be ‘true’ signal for each RNA-Seq sample (Figure 10). The top ten samples with the highest average posterior probability are picked as background and compared with patient samples. Then, DEseq is used for differential expression gene identification. As shown in Figure 11A, the green dots represent significant differential expression gene (p-adj < 0.01) between patient and samples from database. Figure 11B shows top 10 abnormal genes and their geneID in patients.\n\nHist plot shows the distribution of gene expression level for gene ‘CELSR2’ in 1671 different brain RNA-Seq samples. The Gaussian mixture model is fitted by the EM algorithm and the noise is filtered out by posterior probability bigger than 0.5.\n\nA) Scatter plot shows significant differential genes (green dot, p-adj < 0.01). B) Boxplot shows top 10 abnormal genes in simulation compared with data from database.\n\n(‘Neuroblastoma’ related genes is used here).\n\nThe code was tested using the related individuals from the 1000 genomes project. Flagging the genes most likely to have deleterious alleles decreased the search space enough to allow the APRIORI algorithm to run on the dataset.\n\n\nConclusion and next steps\n\nCommon questions in the community about hackathons include whether they can focus on specific diseases and how clinical personnel can interact more effectively with data scientists. We found that it was indeed possible to focus on a given disease while developing generalized tools in a hackathon. In fact, we found it helpful to have cases to use in our analyses from a specific disease. Finally, we found it was to the benefit of everyone to have clinical personnel involved, especially in the later stages of the event.\n\n\nSoftware availability\n\nSource code: https://github.com/NCBI-Hackathons/UPWARD\n\nArchived source code: http://doi.org/10.5281/zenodo.3236567102\n\nLicense: MIT\n\nSource code: https://github.com/NCBI-Hackathons/Rapid_Clinical_Diagnostics\n\nArchived source code: http://doi.org/10.5281/zenodo.3236563104\n\nLicense: MIT\n\nSource code: https://github.com/NCBI-Hackathons/MassiveSeq/\n\nArchived source code: http://doi.org/10.5281/zenodo.3236565106\n\nLicense: MIT\n\nSource code: https://github.com/NCBI-Hackathons/Phenogeno_Viz\n\nArchived source code: http://doi.org/10.5281/zenodo.3236561111\n\nLicense: MIT\n\nSource code: https://github.com/NCBI-Hackathons/pheno_geno_ataxia\n\nArchived source code: http://doi.org/10.5281/zenodo.3236569112\n\nLicense: MIT\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.", "appendix": "Grant information\n\nRHYJ and GCF are partly supported by iron- and heme-related research funding from American Cancer Society Institutional Research Grant (ACS IRG) [ACS-IRG-14-189-19] and WHC fund [310033]. This IronHack and IronBond project is supported by a National Science Foundation cloud computing platform JetStream award [MCB180202] to RHYJ. The “Iron Hack” event is sponsored by USF genomics. This work was funded by the Intramural Research Program of the National Library of Medicine.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe would like to thank John Adams (University of South Florida, USA), Christian Bréchot (University of South Florida, USA), Robert Desnick (Mount Sinai School of Medicine, USA), Jennifer Farmer (Friedreich’s Ataxia Research Alliance, USA), John Phillips (University of Utah, USA), Hervé Puy (National Institute of Health and Medical Research, Paris VII University, University Paris Diderot, France), Kristen Wheeden (American Porphyria Foundation, USA), Derek Wildman (University of South Florida, USA), and Theresa Zesiewicz (University of South Florida, USA) for the stimulating presentations and questions that were fundamental to our work. We additionally thank Paige Hunt, the USF Genomics Program and USF Omics Hub for extensive logistical support in organizing Iron Hack.\n\nNCBI/NIH/USF biohackathon Iron Hack team\n\nGloria C. Ferreira, Jenna Oberstaller, Renée Fonseca, Justin Gibbons, Thomas E. Keller, Chengqi Wang, Xiaoming Liu, Chang Li, Minh Pham, Guy W. Dayhoff II, Linh M. Duong, Swamy Rakesh Adapa, Luis Tañón Reyes, Luciano Enrique Laratelli, Douglas Franz, Segun Fatumo, ATM Golam Bari, Audrey Freischel, Lindsey Fiedler, Omkar Dokur, Krishna Sharma and Deborah Cragun.\n\n\nReferences\n\nCook A, Giunti P: Friedreich’s ataxia: clinical features, pathogenesis and management. Br Med Bull. 2017; 124(1): 19–30. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBalwani M, Desnick RJ: The porphyrias: advances in diagnosis and treatment. Blood. 2012; 120(23): 4496–4504. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKarim Z, Lyoumi S, Nicolas G, et al.: Porphyrias: A 2015 update. Clin Res Hepatol Gastroenterol. 2015; 39(4): 412–425. PubMed Abstract \| Publisher Full Text\n\nRichter T, Nestler-Parr S, Babela R, et al.: Rare Disease Terminology and Definitions-A Systematic Global Review: Report of the ISPOR Rare Disease Special Interest Group. Value Health. 2015; 18(6): 906–914. PubMed Abstract \| Publisher Full Text\n\nBissell DM, Anderson KE, Bonkovsky HL: Porphyria. N Engl J Med. 2017; 377(9): 862–872. PubMed Abstract \| Publisher Full Text\n\nYasuda M, Chen B, Desnick RJ: Recent advances on porphyria genetics: Inheritance, penetrance & molecular heterogeneity, including new modifying/causative genes. Mol Genet Metab. 2018; pii: S1096-7192(18)30645-0. PubMed Abstract \| Publisher Full Text\n\nLecha M, Puy H, Deybach JC: Erythropoietic protoporphyria. Orphanet J Rare Dis. 2009; 4: 19. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nManceau H, Gouya L, Puy H: Acute hepatic and erythropoietic porphyrias: from ALA synthases 1 and 2 to new molecular bases and treatments. Curr Opin Hematol. 2017; 24(3): 198–207. PubMed Abstract \| Publisher Full Text\n\nPuy H, Gouya L, Deybach JC: Porphyrias. Lancet. 2010; 375(9718): 924–937. PubMed Abstract \| Publisher Full Text\n\nRamanujam VM, Anderson KE: Porphyria Diagnostics-Part 1: A Brief Overview of the Porphyrias. Curr Protoc Hum Genet. 2015; 86: 17.20.1–26. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAnderson KE: Clinical and Laboratory Diagnosis of the Porphyrias. In Handbook of Porphyrin Science (Volume 29) With Applications to Chemistry, Physics, Materials Science, Engineering, Biology and Medicine—Volume 29: Porphyrias and Sideroblastic Anemias. (World Scientific). 2014; 369–414. Publisher Full Text\n\nBarman-Aksözen J, C Wiek P, Bansode VB, et al.: Modeling the ferrochelatase c.315-48C modifier mutation for erythropoietic protoporphyria (EPP) in mice. Dis Model Mech. 2017; 10(3): 225–233. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMinder EI, Schneider-Yin X, Minder CE: Patient-recorded outcome to assess therapeutic efficacy in protoporphyria-induced dermal phototoxicity: a proposal. Health Qual Life Outcomes. 2010; 8: 60. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLangendonk JG, Balwani M, Anderson KE, et al.: Afamelanotide for Erythropoietic Protoporphyria. N Engl J Med. 2015; 373(1): 48–59. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLane AM, McKay JT, Bonkovsky HL: Advances in the management of erythropoietic protoporphyria - role of afamelanotide. Appl Clin Genet. 2016; 9: 179–189. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSardh E, Harper P, Balwani M, et al.: Phase 1 Trial of an RNA Interference Therapy for Acute Intermittent Porphyria. N Engl J Med. 2019; 380(6): 549–558. PubMed Abstract \| Publisher Full Text\n\nBerraondo P, Martini PGV, Avila MA, et al.: Messenger RNA therapy for rare genetic metabolic diseases. Gut. 2019; 68(7): 1323–1330. PubMed Abstract \| Publisher Full Text\n\nBalwani M, Naik H, Anderson KE, et al.: Clinical, Biochemical, and Genetic Characterization of North American Patients With Erythropoietic Protoporphyria and X-linked Protoporphyria. JAMA Dermatol. 2017; 153(8): 789–796. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBonkovsky HL, Maddukuri VC, Yazici C, et al.: Acute porphyrias in the USA: features of 108 subjects from porphyrias consortium. Am J Med. 2014; 127(12): 1233–1241. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTo-Figueras J, Ducamp S, Clayton J, et al.: ALAS2 acts as a modifier gene in patients with congenital erythropoietic porphyria. Blood. 2011; 118(6): 1443–1451. PubMed Abstract \| Publisher Full Text\n\nO’Malley R, Rao G, Stein P, et al.: Porphyria: often discussed but too often missed. Pract Neurol. 2018; 18(5): 352–358. PubMed Abstract \| Publisher Full Text\n\nJaramillo-Calle DA: Porphyria. N Engl J Med. 2017; 377(21): 2100–2101. PubMed Abstract \| Publisher Full Text\n\nBadminton MN, Elder GH, Whatley SD: Clinical and molecular epidemiology of the porphyrias. In Handbook of Porphyrin Science (Volume 29) With Applications to Chemistry, Physics, Materials Science, Engineering, Biology and Medicine—Volume 29: Porphyrias and Sideroblastic Anemias. (World Scientific). 2014; 119–150. Publisher Full Text\n\nGenetics Home Reference: Porphyria. Genetics Home Reference. (Accessed: 5th April 2019). Reference Source\n\nPandolfo M: Friedreich ataxia: the clinical picture. J Neurol. 2009; 256 Suppl 1: 3–8. PubMed Abstract \| Publisher Full Text\n\nVaubel RA, Isaya G: Iron-sulfur cluster synthesis, iron homeostasis and oxidative stress in Friedreich ataxia. Mol Cell Neurosci. 2013; 55: 50–61. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTélot L, Rousseau E, Lesuisse E, et al.: Quantitative proteomics in Friedreich’s ataxia B-lymphocytes: A valuable approach to decipher the biochemical events responsible for pathogenesis. Biochim Biophys Acta Mol Basis Dis. 2018; 1864(4 Pt A): 997–1009. PubMed Abstract \| Publisher Full Text\n\nSelak MA, Lyver E, Micklow E, et al.: Blood cells from Friedreich ataxia patients harbor frataxin deficiency without a loss of mitochondrial function. Mitochondrion. 2011; 11(2): 342–350. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBürk K: Friedreich Ataxia: current status and future prospects. Cerebellum Ataxias. 2017; 4: 4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBulteau AL, Dancis A, Gareil M, et al.: Oxidative stress and protease dysfunction in the yeast model of Friedreich ataxia. Free Radic Biol Med. 2007; 42(10): 1561–1570. PubMed Abstract \| Publisher Full Text\n\nNichol H, Gakh O, O'Neill HA, et al.: Structure of frataxin iron cores: an X-ray absorption spectroscopic study. Biochemistry. 2003; 42(20): 5971–5976. PubMed Abstract \| Publisher Full Text\n\nPastore A, Puccio H: Frataxin: a protein in search for a function. J Neurochem. 2013; 126 Suppl 1: 43–52. PubMed Abstract \| Publisher Full Text\n\nChiang S, Kovacevic Z, Sahni S, et al.: Frataxin and the molecular mechanism of mitochondrial iron-loading in Friedreich’s ataxia. Clin Sci (Lond). 2016; 130(11): 853–870. PubMed Abstract \| Publisher Full Text\n\nBencze KZ, Kondapalli KC, Cook JD, et al.: The structure and function of frataxin. Crit Rev Biochem Mol Biol. 2006; 41(5): 269–291. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStehling O, Elsässer HP, Brückel B, et al.: Iron-sulfur protein maturation in human cells: evidence for a function of frataxin. Hum Mol Genet. 2004; 13(23): 3007–3015. PubMed Abstract \| Publisher Full Text\n\nGakh O, Ranatunga W, Smith DY 4th, et al.: Architecture of the Human Mitochondrial Iron-Sulfur Cluster Assembly Machinery. J Biol Chem. 2016; 291(40): 21296–21321. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYoon T, Cowan JA: Frataxin-mediated iron delivery to ferrochelatase in the final step of heme biosynthesis. J Biol Chem. 2004; 279(25): 25943–25946. PubMed Abstract \| Publisher Full Text\n\nMielcarek A, Blauenburg B, Miethke M, et al.: Molecular insights into frataxin-mediated iron supply for heme biosynthesis in Bacillus subtilis. PLoS One. 2015; 10(3): e0122538. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLesuisse E, Santos R, Matzanke BF, et al.: Iron use for haeme synthesis is under control of the yeast frataxin homologue (Yfh1). Hum Mol Genet. 2003; 12(8): 879–889. PubMed Abstract \| Publisher Full Text\n\nBulteau AL, O'Neill HA, Kennedy MC, et al.: Frataxin acts as an iron chaperone protein to modulate mitochondrial aconitase activity. Science. 2004; 305(5681): 242–245. PubMed Abstract \| Publisher Full Text\n\nTamarit J, Obis È, Ros J: Oxidative stress and altered lipid metabolism in Friedreich ataxia. Free Radic Biol Med. 2016; 100: 138–146. PubMed Abstract \| Publisher Full Text\n\nYe H, Rouault TA: Human iron-sulfur cluster assembly, cellular iron homeostasis, and disease. Biochemistry. 2010; 49(24): 4945–4956. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRouault TA, Tong WH: Iron-sulfur cluster biogenesis and human disease. Trends Genet. 2008; 24(8): 398–407. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFerreira GC: Handbook of Porphyrin Science: with Applications to Chemistry, Physics, Materials Science, Engineering, Biology and Medicine - Volume 29: Porphyrias and Sideroblastic Anemias. (World Scientific Publishing Company Pte Limited). 2013. Publisher Full Text\n\nHe Y, Alam SL, Proteasa SV, et al.: Yeast frataxin solution structure, iron binding, and ferrochelatase interaction. Biochemistry. 2004; 43(51): 16254–16262. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSöderberg C, Gillam ME, Ahlgren EC, et al.: The Structure of the Complex between Yeast Frataxin and Ferrochelatase: CHARACTERIZATION AND PRE-STEADY STATE REACTION OF FERROUS IRON DELIVERY AND HEME SYNTHESIS. J Biol Chem. 2016; 291(22): 11887–11898. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAranca TV, Jones TM, Shaw JD, et al.: Emerging therapies in Friedreich’s ataxia. Neurodegener Dis Manag. 2016; 6(1): 49–65. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPineda M, Arpa J, Montero R, et al.: Idebenone treatment in paediatric and adult patients with Friedreich ataxia: long-term follow-up. Eur J Paediatr Neurol. 2008; 12(6): 470–475. PubMed Abstract \| Publisher Full Text\n\nLi L, Voullaire L, Sandi C, et al.: Pharmacological screening using an FXN-EGFP cellular genomic reporter assay for the therapy of Friedreich ataxia. PLoS One. 2013; 8(2): e55940. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchöls L, Zange J, Abele M, et al.: L-carnitine and creatine in Friedreich’s ataxia. A randomized, placebo-controlled crossover trial. J Neural Transm (Vienna). 2005; 112(6): 789–796. PubMed Abstract \| Publisher Full Text\n\nPandolfo M, Arpa J, Delatycki MB, et al.: Deferiprone in Friedreich ataxia: a 6-month randomized controlled trial. Ann Neurol. 2014; 76(4): 509–521. PubMed Abstract \| Publisher Full Text\n\nTomassini B, Arcuri G, Fortuni S, et al.: Interferon gamma upregulates frataxin and corrects the functional deficits in a Friedreich ataxia model. Hum Mol Genet. 2012; 21(13): 2855–2861. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGottesfeld JM, Rusche JR, Pandolfo M: Increasing frataxin gene expression with histone deacetylase inhibitors as a therapeutic approach for Friedreich’s ataxia. J Neurochem. 2013; 126 Suppl 1: 147–154. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHerman D, Jenssen K, Burnett R, et al.: Histone deacetylase inhibitors reverse gene silencing in Friedreich’s ataxia. Nat Chem Biol. 2006; 2(10): 551–558. PubMed Abstract \| Publisher Full Text\n\nLynch DR, Hauser L, McCormick A, et al.: Randomized, double-blind, placebo-controlled study of interferon-γ 1b in Friedreich Ataxia. Ann Clin Transl Neurol. 2019; 6(3): 546–553. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLynch DR, McCormick A, Schadt K, et al.: Pediatric Ataxia: Focus on Chronic Disorders. Semin Pediatr Neurol. 2018; 25: 54–64. PubMed Abstract \| Publisher Full Text\n\nLi Y, Polak U, Clark AD, et al.: Establishment and Maintenance of Primary Fibroblast Repositories for Rare Diseases-Friedreich’s Ataxia Example. Biopreserv Biobank. 2016; 14(4): 324–329. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChen B, Solis-Villa C, Hakenberg J, et al.: Acute Intermittent Porphyria: Predicted Pathogenicity of HMBS Variants Indicates Extremely Low Penetrance of the Autosomal Dominant Disease. Hum Mutat. 2016; 37(11): 1215–1222. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKaplan S, Itzkovitz S, Shapiro E: A universal mechanism ties genotype to phenotype in trinucleotide diseases. PLoS Comput Biol. 2007; 3(11): e235. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLynch DR, Pandolfo M, Schulz JB, et al.: Common data elements for clinical research in Friedreich’s ataxia. Mov Disord. 2013; 28(2): 190–195. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBabady NE, Carelle N, Wells RD, et al.: Advancements in the pathophysiology of Friedreich’s Ataxia and new prospects for treatments. Mol Genet Metab. 2007; 92(1–2): 23–35. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPatel PI, Isaya G: Friedreich ataxia: from GAA triplet-repeat expansion to frataxin deficiency. Am J Hum Genet. 2001; 69(1): 15–24. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMarian AJ, van Rooij E, Roberts R: Genetics and Genomics of Single-Gene Cardiovascular Diseases: Common Hereditary Cardiomyopathies as Prototypes of Single-Gene Disorders. J Am Coll Cardiol. 2016; 68(25): 2831–2849. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSassa S: Gene-environmental interactions: Lessons from porphyria. Environ Health Prev Med. 2003; 7(6): 254–263. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAnderson KE, Bloomer JR, Bonkovsky HL, et al.: Recommendations for the diagnosis and treatment of the acute porphyrias. Ann Intern Med. 2005; 142(6): 439–450. PubMed Abstract \| Publisher Full Text\n\nLiu X, Wu C, Li C, et al.: dbNSFP v3.0: A One-Stop Database of Functional Predictions and Annotations for Human Nonsynonymous and Splice-Site SNVs. Hum Mutat. 2016; 37(3): 235–241. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiu X, White S, Peng B, et al.: WGSA: an annotation pipeline for human genome sequencing studies. J Med Genet. 2016; 53(2): 111–112. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYang Y, Muzny DM, Reid JG, et al.: Clinical whole-exome sequencing for the diagnosis of mendelian disorders. N Engl J Med. 2013; 369(16): 1502–1511. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHrdlickova R, Toloue M, Tian B: RNA-Seq methods for transcriptome analysis. Wiley Interdiscip Rev RNA. 2017; 8(1): e1364. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSanchez N, Chapdelaine P, Rousseau J, et al.: Characterization of frataxin gene network in Friedreich’s ataxia fibroblasts using the RNA-Seq technique. Mitochondrion. 2016; 30: 59–66. PubMed Abstract \| Publisher Full Text\n\nSeco-Cervera M, González-Rodríguez D, Ibáñez-Cabellos JS, et al.: Small RNA-seq analysis of circulating miRNAs to identify phenotypic variability in Friedreich’s ataxia patients. Sci Data. 2018; 5: 180021. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nButler JS, Napierala M: Friedreich’s ataxia--a case of aberrant transcription termination? Transcription. 2015; 6(2): 33–36. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNapierala JS, Li Y, Lu Y, et al.: Comprehensive analysis of gene expression patterns in Friedreich’s ataxia fibroblasts by RNA sequencing reveals altered levels of protein synthesis factors and solute carriers. Dis Model Mech. 2017; 10(11): 1353–1369. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZiemann M, Kaspi A, El-Osta A: Digital Expression Explorer 2: a repository of 4.5 trillion uniformly processed RNA-seq reads and counting. zenodo. 2018. Publisher Full Text\n\nelvers. (Github).\n\nSoneson C: ARMOR. (Github).\n\nKim D, Langmead B, Salzberg SL: HISAT: a fast spliced aligner with low memory requirements. Nat Methods. 2015; 12(4): 357–360. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPertea M, Pertea GM, Antonescu CM, et al.: StringTie enables improved reconstruction of a transcriptome from RNA-seq reads. Nat Biotechnol. 2015; 33(3): 290–295. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGTEx Portal. (Accessed: 5th April 2019). Reference Source\n\nAnders S, Huber W: Differential expression analysis for sequence count data. Genome Biol. 2010; 11(10): R106. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPetrovski S, Wang Q, Heinzen EL, et al.: Genic intolerance to functional variation and the interpretation of personal genomes. PLoS Genet. 2013; 9(8): e1003709. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAdzhubei I, Jordan DM, Sunyaev SR: Predicting functional effect of human missense mutations using PolyPhen-2. Curr Protoc Hum Genet. 2013; Chapter 7: Unit7.20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYang H, Wang K: Genomic variant annotation and prioritization with ANNOVAR and wANNOVAR. Nat Protoc. 2015; 10(10): 1556–1566. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAgrawal R, Mannila H, Srikant R, et al.: Fast discovery of association rules. Advances in Knowledge Discovery and Data Mining. 1996. Reference Source\n\n23andMe Genotyping Services for Research. 23andMe for Scientists. (Accessed: 8th April 2019). Reference Source\n\nAutosomal DNA testing comparison chart - ISOGG Wiki. (Accessed: 8th April 2019). Reference Source\n\nSchulz WL, Nelson BG, Felker DK, et al.: Evaluation of relational and NoSQL database architectures to manage genomic annotations. J Biomed Inform. 2016; 64: 288–295. PubMed Abstract \| Publisher Full Text\n\nMessaoudi C, Fissoune R, Badir H: A performance evaluation of NoSQL databases to manage proteomics data. Int J Data Min Bioinform. 2018; 21(1): 70–89. Publisher Full Text\n\nOffice for Civil Rights (OCR): Summary of the HIPAA Security Rule. HHS.gov. 2013; (Accessed: 8th April 2019). Reference Source\n\nHIPAA Compliance Checklist. HIPAA Journal. (Accessed: 8th April 2019). Reference Source\n\nKöster J, Rahmann S: Snakemake--a scalable bioinformatics workflow engine. Bioinformatics. 2012; 28(19): 2520–2522. PubMed Abstract \| Publisher Full Text\n\nLiao Y, Smyth GK, Shi W: The Subread aligner: fast, accurate and scalable read mapping by seed-and-vote. Nucleic Acids Res. 2013; 41(10): e108. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPertea M, Kim D, Pertea GM, et al.: Transcript-level expression analysis of RNA-seq experiments with HISAT, StringTie and Ballgown. Nat Protoc. 2016; 11(9): 1650–1667. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTsuyuzaki K, Nikaido I: metaSeq: Meta-analysis of RNA-Seq count data in multiple studies. R Package. version 1, 2013. Reference Source\n\nTarazona S, Furió-Tarí P, Turrà D, et al.: Data quality aware analysis of differential expression in RNA-seq with NOISeq R/Bioc package. Nucleic Acids Res. 2015; 43(21): e140. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHänzelmann S, Castelo R, Guinney J: GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics. 2013; 14: 7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiberzon A, Birger C, Thorvaldsdóttir H, et al.: The Molecular Signatures Database (MSigDB) hallmark gene set collection. Cell Syst. 2015; 1(6): 417–425. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nfast.ai · Making neural nets uncool again. (Accessed: 9th April 2019). Reference Source\n\nMikolov T, Chen K, Corrado G, et al.: Efficient Estimation of Word Representations in Vector Space. arXiv [cs.CL]. 2013. Reference Source\n\nGeraldine_VdAuwera: The GATK Best Practices for variant calling on RNAseq, in full detail. GATK-Forum. 2014; (Accessed: 9th April 2019). Reference Source\n\nUPWARD. (Github). Reference Source\n\nFonseca R, Pham M, luistanonreyes, et al.: NCBI-Hackathons/UPWARD v1.0.0. Zenodo. 2019. http://www.doi.org/10.5281/zenodo.3236567\n\nRapid_Clinical_Diagnostics. (Github). Reference Source\n\nChang-Li, Busby B: NCBI-Hackathons/Rapid_Clinical_Diagnostics v1.0.0. Zenodo. 2019. http://www.doi.org/10.5281/zenodo.3236563\n\nMassiveSeq. (Github). Reference Source\n\nCancerGenetics, Keller T, Franz DM, et al.: NCBI-Hackathons/MassiveSeq v1.0.0. 2019. http://www.doi.org/10.5281/zenodo.3236565\n\nVetter T: install-ruby-and-rails. (Github). Reference Source\n\nDownload Python: Python.org. (Accessed: 5th April 2019). Reference Source\n\nRipley BD: The R project in statistical computing. MSOR Connections. The newsletter of the LTSN Maths, Stats & OR Network. 2001; 1: 23–25.\n\nPhenogeno_Viz. (Github). Reference Source\n\nlfiedlerc, DokurOmkar, Busby B: NCBI-Hackathons/Phenogeno_Viz v1.0.0. Zenodo. 2019. http://www.doi.org/10.5281/zenodo.3236561\n\nL, Gibbons J, oberstal, et al.: NCBI-Hackathons/pheno_geno_ataxia v1.0.0. Zenodo. 2019. http://www.doi.org/10.5281/zenodo.3236569\n\npheno_geno_ataxia. (Github). Reference Source" }	[ { "id": "51424", "date": "28 Aug 2019", "name": "Karl E Anderson", "expertise": [ "Reviewer Expertise Clinical and biochemical aspects of the porphyrias" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\n“Iron Hack” is a rare disease-focused hackathon. As a collaborative, problem-solving effort it has successfully attracted participants of diverse backgrounds at the U of South Florida and focusses on rare iron-related genetic diseases, specifically porphyrias and Friedreich’s ataxia.\n\nIt is pitched at a general audience, but some of the material is probably too technical for a general audience. So any revisions should aim to further simplify and explain areas of discussion that are complex and technical.\n\nIntroduction: Friedreich’s ataxia – It would be useful to add a paragraph on diagnosis, and the underlying genetic cause.\n\nThe proposed projects are interesting and timely. These are (1) exploration of consumer-genomics data, (2) largescale RNAseq data mining, (3) genomic data visualization, (4) rare-disease variants discovery, and (5) genotype-to-phenotype mapping. These all seem innovative. The scope of work is very large, and it might be asked whether engaging a larger working group from many institutions would lead to more rapid progress.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] }, { "id": "56584", "date": "04 Dec 2019", "name": "Robert J. Desnick", "expertise": [ "Reviewer Expertise Medical Geneticist with specific expertise in inherited metabolic diseases", "including the clinical", "biochemical", "and molecular aspects of the porphyrias" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nGeneral Comments:\nThis is an interesting and insightful manuscript that describes several novel computer-based algorithms that can investigate and/or uncover new insights into the porphyrias and Fredrich Afaxia. It proves the usefulness of bringing together experts in various fields as a team to stove certain scientific problems. The Iron Hack was very successful and in a short time accomplished an amazing amount of work.\n\nSpecific Comments:\nMaturation and implementation of the various projects may prove the significance of their newly construction programs. Relevant to the porphyrias, it is important that UPWARD does not mislead symptomatic individuals to use consumer/commercial chips (e.g., 23andMe; Ancestry) to establish a diagnosis. This approach has already misguided people to believe they have a porphyria; However, they do not have the “gold standard” biochemical evidence of a specific porphyria. In fact the chip SNP data can lead to individuals having more than one porphyria based on being heterozygous in SNPs for two or more heme biosynthetic genes that are common and benign.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-1135
https://f1000research.com/articles/8-1133/v1	19 Jul 19	{ "type": "Case Report", "title": "Case Report: Penile necrosis associated to paraphimosis with calciphylaxis due to terminal chronic kidney disease", "authors": [ "J. Antonio Grandez-Urbina", "Elizabeth Corrales-Acosta", "J. Eduardo Tejeda-Mariaca", "Rafael Pichardo-Rodriguez", "Herney Garcia-Perdomo", "Elizabeth Corrales-Acosta", "J. Eduardo Tejeda-Mariaca", "Rafael Pichardo-Rodriguez", "Herney Garcia-Perdomo" ], "abstract": "Background: Penile necrosis is a rare condition that may present in patients with diabetes mellitus or chronic kidney disease (CKD). The recommended treatment is controversial. We report a case of penile necrosis in a diabetic patient caused by episode of paraphimosis associated with uremic arteriopathy treated with partial amputation. Clinical Case: A 53-year-old male with a background of arterial hypertension, diabetes mellitus and CKD in hemodialysis. The patient presented with paraphimosis and glans necrosis. An emergency circumcision was carried out. A doppler ultrasound found fluid collection in the left corpus cavernosum, parietal vascular calcifications and vascular insufficiency in the corpus cavernosum that suggested necrosis. A partial amputation of the penis was carried out. After three years of follow-up, the outcome has remained favorable. Conclusions: Penile necrosis is a rare but serious complication of terminal CKD. In these patients, systemic calciphylaxis is usually observed. The main take-away lesson is that a multidisciplinary approach is necessary.", "keywords": [ "Penile necrosis", "Calciphylaxis", "Nephropathy" ], "content": "Introduction\n\nCalciphylaxis is the process of calcification in small and medium vessels, resulting in necrosis in distal regions of the body such as the lower extremities and the penis, the latter being very infrequent1. This condition may present in patients with diabetes mellitus and/or chronic kidney disease (CKD). It has an incidence of 1–4% in patients with CKD that receive hemodialysis2. Due to the technical advances in hemodialysis and the large number of patients that receive this treatment for long periods of time, the number of cases and risk of calciphylaxis has grown3. Calcific uremic arteriolopathy (CUA), also known as calciphylaxis, is a rare complication of CKD where there is occlusion of microvasculature with mural calcification of the arterioles, causing severe ischemia and necrosis of the tissue4. CUA is a major complication of CKD, which demands timely diagnosis. It carries a high risk of mortality and various complications4.\n\nIt is possible to develop ischemic disorders due to arterial calcifications produced by hypercalcemia, as a result of the hyperparathyroidism secondary to CKD5. There are different therapeutic options, but treatment is still controversial. There are no well-established treatment protocols for penile calciphylaxis, and most regimens have been shown to have modest success at best. Treatment options include local wound care, partial or total penectomy, parathyroidectomy, sodium thiosulfate and revascularization6,7.\n\nOn the other hand, paraphimosis occurs when the foreskin of the penis is retracted over the glans and cannot be replaced in its normal position. The tight ring of preputial skin constricts the distal penis causing vascular occlusion and, if not dealt with quickly, can lead to tissue necrosis and partial amputation8. Complications are time related most commonly due to misdiagnosis8. There are few publications related to this uncommon complication.\n\nWe report a case of a patient with end stage renal disease (ESRD) that developed calciphylaxis and consequently distal necrosis of the penis and was treated with an amputation, which resulted in a favorable outcome.\n\n\nCase report\n\nA 53-year-old, Mestizo patient that works as an accounting assistant was admitted to our hospital in emergency room. The patient had a prior medical history of arterial hypertension, diabetes mellitus type 2 and ESRD, for which the patient had been undergoing hemodialysis for two years prior to the initial consultation. No prior history of any surgical intervention was indicated. The patient presented to the emergency ward with an ulcerative and painful lesion in the glans, which had been present for two weeks.\n\nA physical examination revealed foul-smelling distal necrosis of the penis and paraphimosis (Figure 1) with non-palpable inguinal nodes. No other relevant physical findings were described. The laboratory examination showed elevated creatinine (6.36 mg/dL), urea (114 mg/dL), glucose (119 mg/dL), elevated potassium (5.06 mmol/L), C-reactive protein (8.23 mg/dL) and seric calcium (10.1 mg /dl). On the other hand, hemoglobin (7.9 mg/dL), sodium (134 mmol/L) and albumin (2.9 g/dL) levels were found to be below the normal range. No other significant abnormalities were noted. Doppler ultrasound scans showed fluid collection in the left cavernous body, parietal vascular calcifications and vascular insufficiency in both cavernous bodies, suggestive of penile necrosis. The pathology report confirmed the diagnosis of ischemic penile necrosis due to systemic calciphylaxis.\n\nThe patient was treated with a standard circumcision and resection of the scar in the emergency operating room. Inflammatory tissue was evident but wound dehiscence was not observed in the postoperative period. Subsequently, he received 30 sessions in the hyperbaric oxygen chamber over four weeks. The treatment was carried out at 2.8 absolute atmospheres for a duration of two hours. Broad antibiotic therapy was used in order to reduce the progression of the necrosis (a corrected dose for patients on hemodialysis of 250g Imipenem three times a day) for two weeks. Despite the treatment, the clinical response was not a favorable; we identified by physical examination that the necrotic lesion continued, and the general state of the patient began to worsen. Therefore, in order to avoid a worsening in the infection it was decided that a partial amputation of the penis with preservation of 4cm of penile length would be performed, which had a good outcome, evidencing clinical improvement and a decrease in C-reactive protein (Figure 2).\n\nAfter three years of follow-up, the patient did not present with urinary symptoms or pain. The patient can void spontaneously. A timeline of the patient’s medical history, interventions and follow-up is shown in Figure 3.\n\n\nDiscussion\n\nPenile necrosis in patients with ESRD due to calciphylaxis is infrequent9. In ESRD, the origin of the calciphylaxis is related to secondary hyperparathyroidism, as a result of chronic hyperphosphatemia10. Calciphylaxis results in the total obstruction of the arterial lumen due to arterial wall calcification, producing acute ischemia and necrosis of the affected tissue3. In our case, the arterial irrigation of the penis was affected, producing distal necrosis, an unusual event that is usually avoided due to abundant irrigation and collateral circulation. It represents a poor prognostic sign in ESRD patients and is an indicator of metastatic vascular calcification. Pathogenesis of this life-threatening condition is not clearly understood, and treatment is also controversial11.\n\nThe absence of follow-up and treatment with calcitriol for the control of calcium metabolism prior to diagnosis, adding to the absence of treatment for secondary hyperparathyroidism and subsequent calciphylaxis, were the main limitations in this case. However, the post-treatment follow-up and the histological confirmation of the calciphylaxis were important strengths that allowed us to confirm that following the guidelines in previous reports, which recommend starting with the conservative management and following with surgical treatment, were effective and safe with a satisfactory conclusion, despite the high mortality reported for this group of patients.\n\nPenile necrosis produced by calciphylaxis is presented more frequently in patients between 40 and 60 years old. It is associated with ESRD and diabetes mellitus in 100% and 76% of cases respectively, of which our patient had both of risk factors3. Other risk factors obesity, arterial hypertension, use of corticoids, use of inhibitors of vascular calcification and oral anticoagulants3. Our patient had a history of arterial hypertension.\n\nThe image studies to perform include doppler ultrasound of the penis, computed tomography (CT) and magnetic resonance imaging (MRI)12. It is suggested that doppler ultrasound is performed as the first line image study to evaluate vascular permeability and blood flow of the penile vessels. CT could be performed secondarily to evaluate with more sensitivity the extension of the vascular calcification into the soft tissues, necrotizing infection of soft tissues and/or ischemia with the presence of air12. MRI scans can identify with more specificity the necrotic borders of the affected tissues12. In our case, we could only perform a doppler ultrasound within the clinical context of the patient. The rest of the image studies could not be performed as it was necessary to rapidly initiate the required treatment.\n\nFor the management of penile necrosis there are different therapeutic options, from conservative management to surgical intervention, according to the necessity of the case. Nevertheless, it has a poor prognosis and management is controversial9,12. In addition to therapy for penile necrosis, it is recommended to initiate treatment for secondary hyperparathyroidism, in order to reduce long term mortality6,13. In patients undergoing hemodialysis with hyperphosphatemia, it is recommended to use phosphate blockers that do not contain calcium, and in patients with elevated levels of parathormone (PTH), it is recommended to use cinacalcet11. Additionally, studies have reported that necrotic tissue can become a culture medium for multiple microorganisms, so the use of broad-spectrum antibiotics is recommended as a prophylactic measure and empirical therapy11. Empirical antibiotic therapy was used for this patient, as well as surgical treatment, due to the poor response to conservative measures. Treatment for hyperphosphatemia could not be initiated because of the unavailability of the drugs. Karpman et al. demonstrated in a case series of 34 patients that survival rate after partial penile amputation and thyroidectomy was superior to partial amputation by itself, with rates of 75% and 28% respectively. The overall mortality was 64%, with a mean time of 2.5 months until death6.\n\nThe patient had a favorable recovery after partial penectomy without recurrence of necrosis, despite the failure of the first therapeutic line, even though it has been reported that an increase of vascular flow can improve oxygen transport to the ischemic site11,12.\n\n\nConclusions\n\nPenile necrosis constitutes a rare but serious complication associated with ESRD.\n\nRapid and timely management of paraphimosis could improve outcomes in patients with multiple co-morbidities.\n\nFor a good clinical outcome, it is necessary to have a high clinical suspicion and to have knowledge of the different elements involved in clinical support.\n\nA multidisciplinary approach is necessary in the management of this complication.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\n\nConsent\n\nWritten informed consent for publication of their clinical details and clinical images were obtained from the patient.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nMalthouse T, Lam W, Brewin J, et al.: Penile lesion in end-stage renal failure - cancer or otherwise?: Calcific uremic arteriolopathy presenting with a penile lesion. Can Urol Assoc J. 2015; 9(3–4): 136–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHafner J, Keusch G, Wahl C, et al.: Calciphylaxis: a syndrome of skin necrosis and acral gangrene in chronic renal failure. Vasa. 1998; 27(3): 137–43. PubMed Abstract\n\nBarthelmes L, Chezhian C, Thomas KJ: Progression to wet gangrene in penile necrosis and calciphylaxis. Int Urol Nephrol. 2002; 34(2): 231–5. PubMed Abstract \| Publisher Full Text\n\nAhmed MM, Zakir A, Ahsraf MF, et al.: Chronic Kidney Disease and Calciphylaxis: A Literature Review. Cureus. 2018; 10(9): e3334. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBuades Fuster JM, Sanchís Cortés P, Perelló Bestard J, et al.: Plant phosphates, phytate and pathological calcifications in chronic kidney disease. Nefrologia. 2017; 37(1): 20–28. PubMed Abstract \| Publisher Full Text\n\nKarpman E, Das S, Kurzrock EA: Penile calciphylaxis: analysis of risk factors and mortality. J Urol. 2003; 169(6): 2206–9. PubMed Abstract \| Publisher Full Text\n\nPalmisano F, Gadda F, Spinelli MG, et al.: Glans penis necrosis following paraphimosis: A rare case with brief literature review. Urol Case Rep. 2018; 16: 57–58. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZillioux J, Geisenhoff A, Gray M: Management of penile calciphylaxis. A Case Study. J Wound Ostomy Continence Nurs. 2018; 45(6): 536–539. PubMed Abstract \| Publisher Full Text\n\nStein M, Anderson C, Ricciardi R, et al.: Penile gangrene associated with chronic renal failure: report of 7 cases and review of the literature. J Urol. 1994; 152(6 Pt 1): 2014–6. PubMed Abstract \| Publisher Full Text\n\nMartínez GM, García JMO, Sámano VL, et al.: Necrosis de pene: Experiencia en el Hospital de Especialidades Centro Médico Nacional Siglo XXI. Bol Col Mex Urol. 2005; 20(1): 10–4. Reference Source\n\nHaider I, Siddugi M, Naji W, et al.: Calciphylaxis leading to penile necrosis. J Pak Med Assoc. 2014; 64(6): 711–3. PubMed Abstract\n\nCampbell RA, Alzweri LM, Sopko NA, et al.: Penile Calciphylaxis: The Use of Radiological Investigations in the Management of a Rare and Challenging Condition. Urol Case Rep. 2017; 13: 113–116. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGrande M, Facchini F, La Rosa M, et al.: [Calciphylaxis and penile necrosis: a case report and review of the literature]. Urologia. 2010; 77(Suppl 16): 47–50. PubMed Abstract" }	[ { "id": "51410", "date": "30 Jul 2019", "name": "Felix Grases", "expertise": [ "Reviewer Expertise Renal lithiasis and pathological calcifications" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe clinical case is well documented and due to its low prevalence it deserves to be indexed. The authors indicate that the patient had a favorable recovery after partial penectomy without recurrence of necrosis.\nWhat has been the follow-up time since the patient had undergone penectomy?\n\nWhat measures will be taken to avoid recurrence?\n\nWhy the treatments for hyperphosphatemia are unavailable?\n\nWhy the use of inhibitors of vascular calcification can be considered as risk factors of calciphylaxis?\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Partly\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Partly\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [] }, { "id": "51408", "date": "06 Aug 2019", "name": "Munis M. Ahmed", "expertise": [ "Reviewer Expertise Chronic Kidney Disease", "Calciphylaxis." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting yet possible explanation for the outcome portrayed. Kindly work on the following fronts to improve the article:\nAscertain the word count according to the journal guidelines for case reports.\n\nThe introduction can be made more concise.\n\nPut the lab values in a tabular form.\n\nRemove extra headings from the case report section.\n\nIn the treatment section, put \"Broad-spectrum antibiotics\" instead of \"Broad antibiotics\".\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1133
https://f1000research.com/articles/8-1130/v1	19 Jul 19	{ "type": "Systematic Review", "title": "Prognosis associated with geometric patterns of left ventricular remodeling: systematic review and network meta-analysis", "authors": [ "Qishi Zheng", "Germaine Loo", "Thu-Thao Le", "Luming Shi", "Edwin Shih-Yen Chan", "Calvin W. L. Chin", "Qishi Zheng", "Germaine Loo", "Thu-Thao Le", "Luming Shi", "Edwin Shih-Yen Chan" ], "abstract": "Background: There are four geometric patterns (normal geometry, concentric remodeling, concentric and eccentric hypertrophy) used to describe cardiac remodeling. Although left ventricular hypertrophy (LVH) is associated with adverse prognosis, the incremental prognostic value of geometric patterns is less certain. We examined characteristics and prognosis associated with the four conventional patterns of left ventricle (LV) remodeling. Methods: A comprehensive literature search was performed on MEDLINE/PubMed, Embase and the Cochrane Library until January 2019. Network meta-analysis was used to pool data from direct and indirect prognostic comparisons of the four geometric patterns. All-cause mortality was defined as the study outcome. Results: A total of 22 echocardiographic studies (76,142 individuals; 50.1% males; 64.4±7.9 years) of diverse cardiovascular diseases were included. Concentric LVH was associated with the highest prevalence of cardiovascular risk factors and diseases; and eccentric hypertrophy was associated with a high prevalence of atrial fibrillation and low LV ejection fraction. Compared to normal geometry, the risk of all-cause mortality was increased in concentric hypertrophy (risk ratio 1.97 [95% confidence interval 1.63-2.39]) but similar to eccentric hypertrophy (risk ratio 1.15 [95% confidence interval 0.97-1.36]). Conclusions: The study populations examined in the meta-analysis were heterogeneous. Concentric LVH conferred the highest risk of all-cause mortality that overlapped with eccentric hypertrophy. Strategies to improve LVH risk stratification should be examined in future research.", "keywords": [ "Left ventricular hypertrophy", "geometric patterns of left ventricular remodeling", "concentric remodeling", "concentric hypertrophy", "eccentric hypertrophy" ], "content": "Introduction\n\nThe left ventricle (LV) remodels as a response to cardiovascular disease and myocardial injury. Characterized by an increase in LV myocardial mass, left ventricular hypertrophy (LVH) is an established predictor of poorer cardiovascular outcomes1.\n\nFour classical geometric patterns of LV remodeling have been defined based on LV mass and relative wall thickness: normal, concentric remodeling, concentric and eccentric hypertrophy2. This convenient approach of characterizing LV remodeling has been studied across various patient populations, including patients with coronary artery disease, aortic stenosis, hypertensive heart disease and community-based general populations3–6. Whilst some studies demonstrated prognostic associations with these patterns of LV remodeling, others have not. Knowledge of remodeling patterns (concentric and eccentric hypertrophy) provided particularly limited incremental prognostic information beyond LVH1,7–10.\n\nIn this study, we aim to conduct a comprehensive systematic review and network meta-analysis to examine the characteristics and prognosis associated with the four conventional geometric patterns of LV remodeling.\n\n\nMethods\n\nA comprehensive literature search was performed on MEDLINE/PubMed (1946 onwards), Embase (1974 onwards) and the Cochrane Library (1996 onwards) until January 2019. Full text publications evaluating the four conventional LV geometry patterns (normal geometry, concentric remodeling, concentric and eccentric hypertrophy) and prognosis were included. The basic search protocol and specific terms used in the search strategy are available as Extended data11. We conducted the literature search using Medical Subject Headings or Emtree, and free text terms. There were no restrictions on language.\n\nTwo investigators (Q.Z. and G.L.) independently searched for eligible studies based on the pre-defined eligibility criteria. Full-text studies that compared the prognosis of the four conventional LV geometry patterns (i.e. normal geometry, concentric remodeling, concentric and eccentric hypertrophy) were included. We excluded publications in non-adult populations, case reports, commentaries, abstracts, letters-to-editors and review articles. The bibliography in the identified publications and review articles were also reviewed.\n\nThe following data were extracted in duplicates by the two investigators (Q.Z. and G.L.) from the included studies: (1) study characteristics (publication year and patient population); (2) baseline characteristics (mean age, sex distribution, and proportion of patients with hypertension, coronary artery disease, diabetes and other significant risk factors); (3) the four LV remodeling patterns; and (4) adverse prognosis defined as all-cause mortality. Eligible studies that did not report all-cause mortality as an end-point were still included to examine clinical characteristics associated with geometric patterns of LV remodeling. Any disagreements were resolved by discussion with a third investigator (C.W.L.C.). In publications with survival curves, the cumulative survival rates were estimated by digitizing the plots (WebPlotDigitizer version 3.9, Austin, Texas, USA).\n\nTwo investigators (C.W.L.C. and Z.Q.) independently appraised the quality of each study using the Quality In Prognosis Studies tool12. Six domains (study participation, study attrition, prognostic factor measurement, outcome measurement, study confounding; and statistical analysis and reporting) were evaluated to assess the risk of bias in the prognostic studies. In each of the six domains, the risk of bias was classified as “low”, “moderate” or “high”.\n\nA network geometry of the four LV remodeling patterns was constructed. Each node represented a remodeling pattern and its size was weighted by the number of individuals in that group. The connecting line between two nodes denoted direct comparison and its thickness reflected the number of studies included.\n\nThe random-effects meta-regression models were used to measure the impact of baseline characteristics on the effect size of the outcome. The risk ratio (RR) of each LV remodeling group was estimated and reported in the study. To rank the prognosis of all the geometric patterns, we used surface under the cumulative ranking (SUCRA) values13. Rank probabilities of all the groups were first estimated, then followed by a step function to summarize the cumulative ranking for estimating the SUCRA values of each group, ranging from 0 to 100%. Larger SUCRA values indicated better prognosis.\n\nBoth node-splitting and inconsistency modeling were used to test the consistency assumption. The former method involved fitting a series of node-splitting models, one model for each group pair in which there was direct and indirect comparisons. In the latter method, an inconsistency model was fitted and the global Wald test would determine if significant inconsistency was present14. Statistical analyses were performed using Stata/MP Version 13 (StataCorp., College Station, Texas, USA), with the network and network graphs package.\n\n\nResults\n\nFrom an initial 257 publications, 22 echocardiographic studies of diverse cardiovascular diseases satisfied inclusion/exclusion criteria and were included in this study (Figure 1)6–10,15–31. The thresholds used to define LVH and increased concentricity were heterogeneous across the studies (Table 1).\n\nThe systematic review and network meta-analysis was conducted according to the guidelines recommended by PRISMA.\n\n* Two populations were studied in the same publication.\n\nRWT regional wall thickness; HFpEF heart failure preserved ejection fraction; EF ejection fraction; TAVI transcatheter aortic valve implantation\n\nOf the 76,142 individuals pooled from the 22 studies (50.1% males; 64.4±7.9 years), 49.7% had normal geometry; and 31.1%, 10.5% and 8.7% had concentric remodeling, concentric and eccentric hypertrophy, respectively. The proportion of females with concentric and eccentric hypertrophy was high (40–45%). Compared to the other geometric patterns, concentric hypertrophy was associated with the highest prevalence of cardio-metabolic risk factors and cardiovascular diseases. Eccentric hypertrophy was associated with a high prevalence of atrial fibrillation and low LV ejection fraction (Table 2).\n\nMost of the studies demonstrated low risk of bias in the six domains examined (Table 3). The network geometry of LV remodeling patterns was constructed in Figure 2. Concentric remodeling was associated with higher all-cause mortality compared to normal geometry (RR 1.56 [95% confidence interval (CI): 1.31 to 1.85]), and a lower mortality risk compared to concentric hypertrophy (RR 0.79 [95% CI 0.67 to 0.93]). The mortality risk of concentric remodeling was similar compared to eccentric hypertrophy (RR 0.91 [95% CI 0.76 to 1.09]) (Table 4).\n\nThe numbers on the connecting lines denote the studies included for direct comparison.\n\nResults presented in risk ratio and corresponding 95% confidence interval.\n\n*p-value < 0.05\n\nCompared to normal geometry, concentric hypertrophy was associated with highest risk of all-cause mortality (RR 1.97 [95% CI 1.63 to 2.39]; Table 4). The confidence limits overlapped with eccentric hypertrophy (RR 1.71 [95% CI 1.43 to 2.04]). Moreover, the mortality risk of concentric hypertrophy was not significantly increased compared to eccentric LVH (RR 1.15 [95% CI 0.97 to 1.36]). Based on the SUCRA values, the geometric patterns ranked from best to worst prognosis were: normal geometry, concentric remodeling, eccentric hypertrophy and concentric hypertrophy (Figure 3).\n\nA, normal geometry; B, concentric remodeling; C, concentric hypertrophy; D, eccentric hypertrophy.\n\nResults from both node-splitting method and inconsistency model showed no evidence on the violation of consistency assumption between direct and indirect comparisons. Specifically, the pooled estimates between models of consistency (red diamonds) and inconsistency (green diamonds) were identical because all the studies included the four remodeling patterns (Figure 4).\n\n(a) Rank probabilities of effectiveness and SUCRA scores; and (b) prognosis of the four geometric patterns of left ventricular hypertrophy.\n\n\nDiscussion\n\nIn this systematic review and network meta-analysis of 22 echocardiographic publications (n=76,133 individuals), we report the characteristics and prognosis associated with the different patterns of LV remodeling. The study populations were heterogeneous and, importantly, the definitions used to classify the geometric patterns were not uniform. Concentric hypertrophy is associated with the highest prevalence of cardiometabolic risk factors and diseases. Eccentric hypertrophy is associated with a high prevalence of atrial fibrillation and the lowest LV ejection fraction. Although concentric hypertrophy is associated with the highest risk of all-cause mortality, the risks overlapped with eccentric hypertrophy. Eccentric hypertrophy has a similar mortality risk compared to concentric remodeling.\n\nThe pathophysiology of LVH has been well described and studied for the past 50 years32,33. Cardiac hypertrophy is initially an adaptive response to the wall stress according to the Law of LaPlace. Ultimately, cardiac decompensation occurs as a consequence of myocyte death and myocardial fibrosis34,35. Whilst geometric patterns of LV remodeling are clinically meaningful to describe the hypertrophic response due to mechanical stress from either pressure (concentric hypertrophy) or volume overload (eccentric hypertrophy), it may not adequately identify the transition point where adaptive hypertrophy decompensates (Figure 5). This transition point before cardiac decompensation occurs is an important potential risk marker to target more intensive management and closer surveillance. In this study, we have demonstrated that both concentric and eccentric hypertrophy were associated with similar risks of increased all-cause mortality. These observations may suggest that the risk of adverse prognosis is increased once LVH develops, regardless of geometric patterns. It may also suggest that some patients with concentric or eccentric LVH may be in the compensated phase and begets the question of whether there are other strategies to identify high-risk LVH phenotypes.\n\nGeometric patterns of left ventricular remodeling are useful to identify the mechanisms of mechanical stress; but may not adequately identify the transition between myocardial adaptation and decompensation.\n\nTo address the complex interaction between LV dilatation and myocardial thickening in the pathophysiology of LVH, several studies have recently examined an expanded four-group LVH classification: dilated/non-dilated concentric hypertrophy and dilated/non-dilated eccentric hypertrophy36–40. In this proposed four-group LVH classification, dilated concentric hypertrophy was associated with the worst prognosis and non-dilated eccentric hypertrophy had the most favourable profile36–39. However, more guidance is needed before this complex classification can be integrated into routine clinical practice. Recently, we have developed the remodeling index (RI), based on a biophysical model of Laplace’s Law. The RI integrates LV volume and myocardial thickening into a single measurement41. We further demonstrated that hypertensive LVH patients with abnormally low RI (suggestive of excessive myocardial thickening relative to LV dilatation) had increased myocardial fibrosis, elevated circulating markers of myocardial injury and wall stress; and in a small number of patients with dilated cardiomyopathy, an abnormally high RI (suggestive of excessive LV dilatation relative to myocardial thickening) was associated with adverse cardiovascular events41. The prognostic value and clinical utility of this index are currently being examined in a large cohort of hypertensive patients (ClinicalTrials.gov identifier: NCT02670031).\n\nThese emerging data support the notion that cardiac remodeling in hypertrophy is heterogeneous and complex; and the conventional geometric patterns of LV remodeling is not adequate to risk-stratify patients with LVH.\n\nThe study populations included in the meta-analysis were heterogeneous. It is possible that the conventional remodeling patterns has incremental prognostic value in certain cardiac conditions. Unfortunately, the limited number of studies precluded stratified analyses to examine the prognostic value of LV geometric patterns in the different cardiovascular conditions. The definitions used for classifying geometric patterns were not consistent across the different studies. This is concerning and reinforces the necessity to apply consensus definitions in future studies2.\n\n\nConclusions\n\nConcentric and eccentric hypertrophy are associated with increased and similar all-cause mortality. Possible explanations for these observations include the heterogeneous populations, inconsistent definitions used in the classification and the inherent limitations of the conventional patterns of LV geometry to adequately risk stratify LVH. Well-validated novel approaches to improve risk stratification of LVH should be explored in future research.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\nOpen Science Framework: Prognosis associated with geometric patterns of left ventricular remodeling: systematic review and network meta-analysis. https://doi.org/10.17605/OSF.IO/3CJMW11.\n\nThis project contains the following extended data:\n\nData file.xlsx (Sheet 1 contains study questions, search date, search terms and eligibility criteria; Sheet 2 contains a list of the studies identified; Sheet 3 contains the six composites used in this study).\n\nOpen Science Framework: PRISMA checklist for “Prognosis associated with geometric patterns of left ventricular remodeling: systematic review and network meta-analysis”. https://doi.org/10.17605/OSF.IO/3CJMW11.\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in funding this work.\n\n\nReferences\n\nVerdecchia P, Porcellati C, Reboldi G, et al.: Left ventricular hypertrophy as an independent predictor of acute cerebrovascular events in essential hypertension. Circulation. 2001; 104(17): 2039–44. PubMed Abstract \| Publisher Full Text\n\nLang RM, Badano LP, Mor-Avi V, et al.: Recommendations for cardiac chamber quantification by echocardiography in adults: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr. 2015; 28(1): 1–39. PubMed Abstract \| Publisher Full Text\n\nLevy D, Garrison RJ, Savage DD, et al.: Prognostic implications of echocardiographically determined left ventricular mass in the Framingham Heart Study. N Engl J Med. 1990; 322(22): 1561–6. PubMed Abstract \| Publisher Full Text\n\nZoccali C, Benedetto FA, Mallamaci F, et al.: Prognostic impact of the indexation of left ventricular mass in patients undergoing dialysis. J Am Soc Nephrol. 2001; 12(12): 2768–74. PubMed Abstract\n\nElliott PM, Gimeno Blanes JR, Mahon NG, et al.: Relation between severity of left-ventricular hypertrophy and prognosis in patients with hypertrophic cardiomyopathy. Lancet. 2001; 357(9254): 420–4. PubMed Abstract \| Publisher Full Text\n\nLieb W, Gona P, Larson MG, et al.: The natural history of left ventricular geometry in the community: clinical correlates and prognostic significance of change in LV geometric pattern. JACC Cardiovasc Imaging. 2014; 7(9): 870–878. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKrumholz HM, Larson M, Levy D: Prognosis of left ventricular geometric patterns in the Framingham Heart Study. J Am Coll Cardiol. 1995; 25(4): 879–84. PubMed Abstract \| Publisher Full Text\n\nGhali JK, Liao Y, Cooper RS: Influence of left ventricular geometric patterns on prognosis in patients with or without coronary artery disease. J Am Coll Cardiol. 1998; 31(7): 1635–40. PubMed Abstract \| Publisher Full Text\n\nKatz DH, Beussink L, Sauer AJ, et al.: Prevalence, clinical characteristics, and outcomes associated with eccentric versus concentric left ventricular hypertrophy in heart failure with preserved ejection fraction. Am J Cardiol. 2013; 112(8): 1158–64. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPaoletti E, De Nicola L, Gabbai FB, et al.: Associations of Left Ventricular Hypertrophy and Geometry with Adverse Outcomes in Patients with CKD and Hypertension. Clin J Am Soc Nephrol. 2015; 11(2): 271–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChin C: Prognosis associated with geometric patterns of left ventricular remodeling: systematic review and network meta-analysis. 2019. http://www.doi.org/10.17605/OSF.IO/3CJMW\n\nHayden JA, van der Windt DA, Cartwright JL, et al.: Assessing bias in studies of prognostic factors. Ann Intern Med. 2013; 158(4): 280–286. PubMed Abstract \| Publisher Full Text\n\nSalanti G, Ades AE, Ioannidis JP: Graphical methods and numerical summaries for presenting results from multiple-treatment meta-analysis: an overview and tutorial. J Clin Epidemiol. 2011; 64(2): 163–71. PubMed Abstract \| Publisher Full Text\n\nWhite IR: Network meta-analysis. Stata J. 2015; 15(4): 951–985. Publisher Full Text\n\nVerma A, Meris A, Skali H, et al.: Prognostic implications of left ventricular mass and geometry following myocardial infarction: the VALIANT (VALsartan In Acute myocardial iNfarcTion) Echocardiographic Study. JACC Cardiovasc Imaging. 2008; 1(5): 582–91. PubMed Abstract \| Publisher Full Text\n\nGardin JM, McClelland R, Kitzman D, et al.: M-mode echocardiographic predictors of six- to seven-year incidence of coronary heart disease, stroke, congestive heart failure, and mortality in an elderly cohort (the Cardiovascular Health Study). Am J Cardiol. 2001; 87(9): 1051–7. PubMed Abstract \| Publisher Full Text\n\nMilani RV, Lavie CJ, Mehra MR, et al.: Left ventricular geometry and survival in patients with normal left ventricular ejection fraction. Am J Cardiol. 2006; 97(7): 959–63. PubMed Abstract \| Publisher Full Text\n\nLavie CJ, Milani RV, Ventura HO, et al.: Left ventricular geometry and mortality in patients >70 years of age with normal ejection fraction. Am J Cardiol. 2006; 98(10): 1396–9. PubMed Abstract \| Publisher Full Text\n\nBerger J, Ren X, Na B, et al.: Relation of concentric remodeling to adverse outcomes in patients with stable coronary artery disease (from the Heart and Soul Study). Am J Cardiol. 2011; 107(11): 1579–84. PubMed Abstract \| Publisher Full Text\n\nVerdecchia P, Schillaci G, Borgioni C, et al.: Prognostic value of left ventricular mass and geometry in systemic hypertension with left ventricular hypertrophy. Am J Cardiol. 1996; 78(2): 197–202. PubMed Abstract \| Publisher Full Text\n\nKohara K, Zhao B, Jiang Y, et al.: Relation of left ventricular hypertrophy and geometry to asymptomatic cerebrovascular damage in essential hypertension. Am J Cardiol. 1999; 83(3): 367–70. PubMed Abstract \| Publisher Full Text\n\nApostolakis S, Sullivan RM, Olshansky B, et al.: Left ventricular geometry and outcomes in patients with atrial fibrillation: the AFFIRM Trial. Int J Cardiol. 2014; 170(3): 303–8. PubMed Abstract \| Publisher Full Text\n\nShah N, Badheka AO, Grover PM, et al.: Influence of left ventricular remodeling on atrial fibrillation recurrence and cardiovascular hospitalizations in patients undergoing rhythm-control therapy. Int J Cardiol. 2014; 174(2): 288–92. PubMed Abstract \| Publisher Full Text\n\nDebry N, Maréchaux S, Rusinaru D, et al.: Prognostic significance of left ventricular concentric remodelling in patients with aortic stenosis. Arch Cardiovasc Dis. 2017; 110(1): 26–34. PubMed Abstract \| Publisher Full Text\n\nShigematsu Y, Hamada M, Ohtsuka T, et al.: Left ventricular geometry as an independent predictor for extracardiac target organ damage in essential hypertension. Am J Hypertens. 1998; 11(10): 1171–7. PubMed Abstract \| Publisher Full Text\n\nGerdts E, Cramariuc D, de Simone G, et al.: Impact of left ventricular geometry on prognosis in hypertensive patients with left ventricular hypertrophy (the LIFE study). Eur J Echocardiogr. 2008; 9(6): 809–15. PubMed Abstract \| Publisher Full Text\n\nFabiani I, Pugliese NR, La Carrubba S, et al.: Incremental prognostic value of a complex left ventricular remodeling classification in asymptomatic for heart failure hypertensive patients. J Am Soc Hypertens. 2017; 11(7): 412–419. PubMed Abstract \| Publisher Full Text\n\nPark CS, Park JB, Kim Y, et al.: Left Ventricular Geometry Determines Prognosis and Reverse J-Shaped Relation Between Blood Pressure and Mortality in Ischemic Stroke Patients. JACC Cardiovasc Imaging. 2018; 11(3): 373–382. PubMed Abstract \| Publisher Full Text\n\nLavie CJ, Milani RV, Patel D, et al.: Disparate effects of obesity and left ventricular geometry on mortality in 8088 elderly patients with preserved systolic function. Postgrad Med. 2009; 121(3): 119–25. PubMed Abstract \| Publisher Full Text\n\nCapoulade R, Clavel MA, Le Ven F, et al.: Impact of left ventricular remodelling patterns on outcomes in patients with aortic stenosis. Eur Heart J Cardiovasc Imaging. 2017; 18(12): 1378–1387. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRymuza B, Zbroński K, Scisło P, et al.: Left ventricular remodelling pattern and its relation to clinical outcomes in patients with severe aortic stenosis treated with transcatheter aortic valve implantation. Postepy Kardiol Interwencyjnej. 2017; 13(4): 288–294. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMeerson FZ: Compensatory hyperfunction of the heart and cardiac insufficiency. Circ Res. 1962; 10: 250–8. PubMed Abstract \| Publisher Full Text\n\nBadeer HS: Biological significance of cardiac hypertrophy. Am J Cardiol. 1964; 14: 133–8. PubMed Abstract \| Publisher Full Text\n\nDiwan A, Dorn GW 2nd: Decompensation of cardiac hypertrophy: cellular mechanisms and novel therapeutic targets. Physiology (Bethesda). 2007; 22: 56–64. PubMed Abstract \| Publisher Full Text\n\nLorell BH, Carabello BA: Left ventricular hypertrophy: pathogenesis, detection, and prognosis. Circulation. 2000; 102(4): 470–9. PubMed Abstract \| Publisher Full Text\n\nKhouri MG, Peshock RM, Ayers CR, et al.: A 4-tiered classification of left ventricular hypertrophy based on left ventricular geometry: the Dallas heart study. Circ Cardiovasc Imaging. 2010; 3(2): 164–71. PubMed Abstract \| Publisher Full Text\n\nBang CN, Gerdts E, Aurigemma GP, et al.: Four-group classification of left ventricular hypertrophy based on ventricular concentricity and dilatation identifies a low-risk subset of eccentric hypertrophy in hypertensive patients. Circ Cardiovasc Imaging. 2014; 7(3): 422–9. PubMed Abstract \| Publisher Full Text\n\nde Simone G, Izzo R, Aurigemma GP, et al.: Cardiovascular risk in relation to a new classification of hypertensive left ventricular geometric abnormalities. J Hypertens. 2015; 33(4): 745–54; discussion 754. PubMed Abstract \| Publisher Full Text\n\nGarg S, de Lemos JA, Ayers C, et al.: Association of a 4-Tiered Classification of LV Hypertrophy With Adverse CV Outcomes in the General Population. JACC Cardiovasc Imaging. 2015; 8(9): 1034–1041. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGaasch WH, Zile MR: Left ventricular structural remodeling in health and disease: with special emphasis on volume, mass, and geometry. J Am Coll Cardiol. 2011; 58(17): 1733–40. PubMed Abstract \| Publisher Full Text\n\nGoh VJ, Le TT, Bryant J, et al.: Novel Index of Maladaptive Myocardial Remodeling in Hypertension. Circ Cardiovasc Imaging. 2017; 10(9): pii: e006840. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "54455", "date": "16 Oct 2019", "name": "Bruno Trimarco", "expertise": [ "Reviewer Expertise Hypertension" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper titled “Prognosis associated with geometric patterns of left ventricular remodeling: systematic review and network meta-analysis” by Zheng et al aims to identify different prognostic profile of different left ventricular remodeling patterns. Although the aim is of interest, the method to select data from the international literature is quite messy. In fact, populations are not homogenous (from population-based studies like the FHS, to patients with stable CAD, to patients with hypertension, aortic stenosis, HFpEF, etc). In addition, methods reported in studies to categorize left ventricular remodeling are still not homogeneous. Consequently, conclusions of the paper are erroneous.\nAuthors stated that the new classification of left ventricular remodeling proposed by Khouri is “complex” and “more guidance is needed before this complex classification can be integrated into routine clinical practice”, so that they propose a more simple index (the remodeling index) which at moment is tested by the authors in a trial of hypertensive patients. We thank the Author for the suggestion, i.e. to use their index; of course, to demonstrate in the future that their index is superior to that used in the past, they will have to perform a correct statistical analysis, confronting their index with clear-cut values suggested by guidelines and more recently by Khouri, (which, I guess, is very easy to use into clinical practice) for the identification of left ventricular remodeling patterns.\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? No\n\nIs the statistical analysis and its interpretation appropriate? No\n\nAre the conclusions drawn adequately supported by the results presented in the review? No", "responses": [ { "c_id": "4975", "date": "21 Oct 2019", "name": "Calvin Chin", "role": "Author Response", "response": "We thank the Reviewer for the comments. We wish to highlight the issues raised by the Reviewer are inherent to the individual studies and not the methodology of the meta analysis. We agree with the Reviewer that the study populations included in the study are heterogeneous. Unfortunately, the limited number of studies precluded further analyses to examine prognostic value of LV geometric patterns in different cardiovascular conditions. The definitions used for classifying geometric patterns were not consistent in the studies as highlighted by the Reviewer, a limitation that reinforces the necessity of applying consensus definitions in future studies. Both of these points have already been listed as study limitations." } ] }, { "id": "55256", "date": "23 Oct 2019", "name": "Tee Joo Yeo", "expertise": [ "Reviewer Expertise Sports cardiology", "Athlete's heart", "physiological vs pathological cardiac remodeling in athletes." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nZheng et al performed a systematic review and meta-analysis of echocardiographic studies that included prognostic characteristics of 4 geometric patterns of cardiac remodeling.\nThrough the combined total of 22 studies comprising diverse populations, including hypertension, aortic stenosis, heart failure with preserved ejection fraction, ischemic strokes, chronic kidney disease etc, the authors conclude that the presence of concentric LVH and eccentric hypertrophy was associated with a higher risk of all-cause mortality.\nThe overall aim of this study is commendable. Nonetheless, there are a few areas of note:\nThe study includes populations both with and without pre-existing cardiovascular disease; in such a situation, the baseline risk profiles differ significantly (eg. an apparently healthy individual with newly diagnosed hypertension has a substantially lower risk profile compared to an individual with pre-existing or longstanding aortic stenosis although both may develop LVH eventually).\n\nA suggested approach would be to stratify the risk of LVH based on aetiology (e.g. hypertension-related, aortic stenosis related, multifactorial etc.) which would make the meta-analysis more applicable to the above sub-groups of patients rather than providing a single risk ratio for an entire phenotypic category of cardiac remodeling.\n\nThe authors may consider excluding the studies involving the general population to limit the study population to those with pre-existing disease, as the absolute number of Framingham subjects with LVH was actually small.\n\nImportantly, the classification of LVH and RWT differed in certain studies, so applicability is affected. Although the authors listed this as a limitation, studies with substantial differences in quantifying LVH might need to be excluded in order to keep the study population homogenous.\n\nIn Table 2, regional (?or relative, as stated by the American Society of Echocardiography and other international echocardiography societies) wall thickness for all 4 categories of cardiac remodeling ranged between 0.95 to 1.29. Do check if this is correct, as that would classify all subjects in the concentric geometry.\n\nFor the conclusion, the authors are right to say that the study populations were heterogeneous. As such, it might be more appropriate to generate different risks for the larger subsets of patients (eg. LVH from hypertension vs LVH from aortic stenosis vs LVH from combined causes) rather than to generalise risk for all patients regardless of aetiology.\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes\n\nIs the statistical analysis and its interpretation appropriate? I cannot comment. A qualified statistician is required.\n\nAre the conclusions drawn adequately supported by the results presented in the review? Partly", "responses": [] } ]	1	https://f1000research.com/articles/8-1130
https://f1000research.com/articles/8-960/v1	24 Jun 19	{ "type": "Research Article", "title": "A genome-wide scan of cleft lip triads identifies parent-of-origin interaction effects between ANK3 and maternal smoking, and between ARHGEF10 and alcohol consumption", "authors": [ "Øystein Ariansen Haaland", "Julia Romanowska", "Miriam Gjerdevik", "Rolv Terje Lie", "Håkon Kristian Gjessing", "Astanand Jugessur", "Julia Romanowska", "Miriam Gjerdevik", "Rolv Terje Lie", "Håkon Kristian Gjessing", "Astanand Jugessur" ], "abstract": "Background: Although both genetic and environmental factors have been reported to influence the risk of isolated cleft lip with or without cleft palate (CL/P), the exact mechanisms behind CL/P are still largely unaccounted for. We recently developed new methods to identify parent-of-origin (PoO) interactions with environmental exposures (PoOxE) and applied them to families with children born with isolated cleft palate only. Here, we used the same genome-wide association study (GWAS) dataset and methodology to screen for PoOxE effects in the larger sample of CL/P triads. Methods: Genotypes from 1594 complete triads and 314 dyads (1908 nuclear families in total) with CL/P were available for the current analyses. Of these families, 1024 were Asian, 825 were European and 59 had other ancestries. After quality control, 341,191 SNPs remained from the original 569,244. The exposures were maternal cigarette smoking, use of alcohol, and use of vitamin supplements in the periconceptional period. The methodology applied in the analyses is implemented in the R-package Haplin. Results: Among Europeans, there was evidence of a PoOxSmoke effect for ANK3 with three SNPs (rs3793861, q=0.20, p=2.6e-6; rs7087489, q=0.20, p=3.1e-6; rs4310561, q=0.67, p=4.0e-5) and a PoOxAlcohol effect for ARHGEF10 with two SNPs (rs2294035, q=0.32, p=2.9e-6; rs4876274, q=0.76, p=1.3e-5). Conclusion: Our results indicate that the detected PoOxE effects have a plausible biological basis, and thus warrant replication in other independent cleft samples. Our demonstration of the feasibility of identifying complex interactions between relevant environmental exposures and PoO effects offers new avenues for future research aimed at unravelling the complex etiology of cleft lip defects.", "keywords": [ "Orofacial cleft", "cleft lip with or without cleft palate", "case-parent triads", "gene-environment interaction", "parent-of-origin", "PoOxE", "Haplin" ], "content": "Introduction\n\nCleft lip with or without cleft palate (CL/P) appears in approximately 3.4 to 22.9 per 10,000 live births1. Based on the severity of the cleft, patients undergo varying degrees of medical, dental, speech and psychosocial interventions over the first two decades of their lives, a long-term multidisciplinary treatment that not only imposes a heavy burden on patients and their families2,3, but also accounts for a substantial outlay in national healthcare budgets4,5.\n\nMultiple genetic and environmental factors have been reported to influence the risk of CL/P, individually and through complex interactions in relevant biological pathways6–10. Major advances in high-throughput genotyping technologies, coupled with a boost in international collaborations, have led to substantial progress in gene-mapping for orofacial clefts, and the first wave of genome-wide association studies (GWAS) identified and replicated several key genes and loci associated with clefting11–16. Despite this success, the genetic variants identified so far collectively explain only a minor fraction of the total variance attributable to additive genetic effects, which is intriguing considering the more than 70% heritability of CL/P among Europeans17–20. This has spurred renewed interest in investigating disease mechanisms other than fetal or maternal effects21. One example is parent-of-origin (PoO), where the effect of a particular allele in the offspring differs according to its parental origin22–24, and another is gene-environment interaction (GxE), where fetal effects differ across strata of environmental exposures25. Identifying GxE effects may not only provide new insights into the causes of CL/P, but may also provide an opportunity to intervene on environmental risk factors alone, particularly in subgroups of the population that are genetically more susceptible to these environmental effects.\n\nRecently, we went one step further and developed new methods for a genome-wide screening for PoO interactions with environmental exposures (i.e., PoOxE) in the case-parent triad setting22. We applied the new methodology, implemented in the R-package Haplin26, to isolated cleft palate only (CPO)27, using genotypes and exposure data from the largest published GWAS dataset on case-parent triads of orofacial clefts11. Epidemiological and embryological findings have previously shown that CL/P and CPO may have distinct etiologies. Therefore, we used the same GWAS dataset and methodology to perform a genome-wide scan for PoOxE effects in the larger sample of isolated CL/P.\n\n\nMethods\n\nThe study participants were mainly of Asian or European origin and were recruited as part of an international cleft collaboration11. Information was available on genotypes as well as maternal vitamin use, cigarette smoking and alcohol consumption in the periconceptional period (three months before and three months after pregnancy). The information on environmental exposures was based on interviews and questionnaires. More detailed characteristics of the study participants can be found in our recent work28.\n\nTable 1 shows the distribution of the CL/P families according to ethnicity, triad completeness and maternal exposure. There were 1908 families in the pooled sample (5424 individuals in total), which included all the participants. Of these, 825 families were in the European sample, 1024 families were in the Asian sample, and 59 families were in the sample consisting of other ethnicities (Table 1). We performed three main sets of analyses on the following samples: All participants (denoted as “pooled analysis”), only Asians (“Asian analysis”), and only Europeans (“European analysis”). The 59 families with other ethnicities were not analyzed due to the small sample size. In the pooled and European analyses, we examined all exposures. As cigarette smoking and alcohol consumption were rare among Asian mothers, we were only able to conduct PoOxVitamin analyses for this ethnicity.\n\naNo analyses of parent-of-origin interactions with alcohol (PoOxAlcohol) or parent-of-origin interactions with smoking (PoOxSmoke) were conducted for this group because of a lack of observations for these exposures.\n\nbOwing to the small sample size, no analysis of parent-of-origin interactions with environmental exposures (PoOxE) was conducted for this group.\n\nNote that a subset of the complete triads included more than one offspring. Incomplete triads are parent-offspring dyads.\n\nQuality control for excluding single-nucleotide polymorphisms (SNPs) and samples were conducted as described in Haaland et al. (2017)27. That is, we included SNPs with a missing call rate less than 5%, a minor allele frequency (MAF) greater than 5%, a p-value of less than 0.001 for the test for Hardy-Weinberg equilibrium presented by Wigginton et al. (2005)29, and a Mendelian error rate greater than 10%. Further, if two or more SNPs were in perfect linkage disequilibrium (r2=1) with each other, we only included one in the analyses. After applying these same criteria here, 341,191 were left for the current analyses from a total of 569,244 SNPs (Table 2).\n\naSome SNPs failed several criteria. Hence, the remaining number of SNPs (341,191) plus the ones that failed the different criteria do not add up to the total number of SNPs (569,244).\n\nHWE, Hardy-Weinberg equilibrium; MAF, minor allele frequency.\n\nFor statistical analysis, we used the statistical software Haplin26, which is written in the R programming language30. Haplin is based on log-linear modeling in a maximum likelihood framework and is well-suited for the analysis of offspring-parent triads. Because Haplin uses the expectation-maximization (EM) algorithm to account for missing parental genotypes26, we were able to include the 314 case-parent dyads in the analyses beside the complete triads (Table 1). Haplin also uses the EM algorithm to reconstruct haplotypes, which enabled haplotype analyses for different combinations of SNPs in the genes that showed a plausible PoOxE effect.\n\nA detailed description of the method for PoOxE analysis has been provided in our previous works22,27,31. Briefly, PoOxE effects were calculated as follows:\n\n1) Calculate the relative risk (RR) for an allele inherited from the mother (RRmat) and do the same for the father (RRpat).\n\n2) Calculate the relative risk ratio (RRRPoO=RRmat/RRpat) between the RRs in (1). RRRPoO is thus an estimate of the parent-of-origin (PoO) effect.\n\n3) Calculate RRRPoOxE as RRRPoO(Exposed)/RRRPoO(Unexposed), where RRRPoO(Exposed) and RRRPoO(Unexposed) are RRRPoO among triads with exposed and unexposed mothers.\n\nHaplin uses a Wald test to test the null hypothesis of RRRPoOxE=1.\n\nIn order to control for multiple testing (one test for each of 341,191 SNPs), we obtained q-values using the false discovery rate (FDR) method described by Storey & Tibshirani (2003)32. Specifically, the q-values were calculated from the p-values with the R-function qvalue()33. A q-value of 0.2 corresponds to an FDR of 20%, which means that at least 80% of SNPs with a q-value less than 0.2 would be expected to be truly associated with the outcome. As in our previous work on isolated CPO27, we identified the top 20 SNPs for each of the analyses performed (see Results for details) and calculated relative risk ratios (RRRs) with 95% confidence intervals (95% CI). We paid more attention to a given gene if SNPs in that gene showed up multiple times in one set or across different sets of analyses.\n\nTo illustrate the general ability of the PoOxE analyses to detect true associations, power analyses for a wide range of PoOxE scenarios were performed using the Haplin function hapPowerAsymp(), as described in our recent works22,34.\n\nWe focused on the regions flanking SNPs in the most interesting genes and constructed regional plots based on R-scripts developed by the Diabetes Genetics Initiative of Broad Institute of Harvard and MIT, Lund University and Novartis Institutes of BioMedical Research35. Such plots capture the extent of linkage disequilibrium between a lead SNP and neighboring SNPs, while also providing information on recombination patterns and the position of genes.\n\nR-scripts used to conduct the statistical analyses and create figures are available (see Software availability)36.\n\nTo contextualize the findings, we searched for connections among a selection of genes in the STRING database37, as well as for enrichment of these genes in expression patterns using ExpressionAtlas38 and BGee (R package BgeeDB_2.10.0)39. Further, using Hetionet (Ver.1.0)40, we searched for indirect links between the genes highlighted by our analyses, the exposures and the phenotype (“cleft lip”). Hetionet is a heterogeneous network of various relationships among various data types, such as interactions between genes, or regulation of gene expression between a drug and a gene. The data used in Hetionet were carefully curated from 29 publicly available databases. To simplify the query output, the number of relationships between any two of the input query nodes (i.e., exposure, cleft lip, and the genes) was set to at most two. The exact queries together with their output are available (see Software availability)36.\n\nThe individual institutional review boards of the members of the International Cleft Consortium provided ethical approval, which can be found in the online supplementary material of the original publication41. Written informed consent was provided by all participating families. Please refer to the dbGaP database for more information.\n\n\nResults\n\nFor clarity, this section is structured as follows: We present the results of the PoOxE analyses of the pooled sample first (Table 3), followed by those of the European (Table 4) and Asian (Table 5) samples. We used the integrative database GeneCards and the gene-centric links therein to collate information on the genes in these tables. The 1000 Genomes browser was used to determine the chromosomal band location of a SNP. In the following sections, we focus on q-values, but all the corresponding p-values can also be found in Table 3–Table 5. Table 6 provides a reference for the full names of all the genes mentioned in Table 3–Table 5. Table 7 shows the results of the haplotype analyses of SNPs in the most interesting genes. Figure 1 and Figure 2 present visualizations of the results from the bioinformatics analyses, and regional plots for the most important regions from Table 3–Table 5 are shown in Figure 3 and Figure 4. Figure 5 illustrates power calculations to detect different PoOxE effects in single-SNP analyses under different parameters, such as different sample sizes and minor allele frequencies. Quantile-quantile (QQ) plots for each set of analyses are shown in Figure 6–Figure 8.\n\naThe 1000 Genomes browser was used to determine the chromosomal band location of a SNP.\n\nbIf a SNP is located within a gene itself, the gene symbol is provided (the full names of the genes are provided in Table 6). SNPs located within 40 kb of a gene have the prefix ‘~’, and those not located within a 40 kb-distance of a gene are denoted as NC (for ‘not close’). Note that pseudogenes and non-coding RNAs are excluded.\n\ncShared: Also in Table 4 or Table 5.\n\nSNP, single-nucleotide polymorphism; RRR, relative risk ratio; CI, confidence interval; NC, not close.\n\naThe 1000 Genomes browser was used to determine the chromosomal band location of a SNP.\n\nbIf a SNP is located within a gene itself, the gene symbol is provided (the full names of the genes are provided in Table 6). SNPs located within 40 kb of a gene have the prefix ‘~’, and those not located within a 40 kb-distance of a gene are denoted as NC (for ‘not close’). Note that pseudogenes and non-coding RNAs are excluded.\n\ncShared: Also in Table 3 or Table 5.\n\nSNP, single-nucleotide polymorphism; RRR, relative risk ratio; CI, confidence interval; NC, not close.\n\naThe 1000 Genomes browser was used to determine the chromosomal band location of a SNP.\n\nbIf a SNP is located within a gene itself, the gene symbol is provided (the full names of the genes are provided in Table 6). SNPs located within 40 kb of a gene have the prefix ‘~’, and those not located within a 40 kb-distance of a gene are denoted as NC (for ‘not close’). Note that pseudogenes and non-coding RNAs (ncRNA) are excluded.\n\nThere is no column for “shared” here, as none of these SNPs featured among those listed in Table 3 or Table 4.\n\nSNP, single-nucleotide polymorphism; RRR, relative risk ratio; CI, confidence interval; NC, not close.\n\naEffect allele or haplotype against the reference. Lowercase indicates the minor allele at the SNP.\n\nbMinor allele frequency for a given SNP. In haplotype analyses, this corresponds to the frequencies of haplotypes other than the reference.\n\ncRR for child effects; RRR for GxSmoke or GxAlcohol, PoO and PoOxSmoke or PoOxAlcohol. All p-values <0.05 are highlighted in bold.\n\nNote that in a two-SNP-haplotype, there are four possible combinations, and in a three-SNP-haplotype there are eight. However, only two or three of these combinations were actually observed in the data.\n\nSNP, single nucleotide polymorphism; RRR, relative risk ratio; RR, risk ratio; CI, confidence interval; PoO, parent-of-origin; GxSmoke, gene-smoking interaction; GxAlcohol, gene-alcohol interaction; PoOxAlcohol, parent-of-origin interactions with alcohol; PoOxSmoke, parent-of-origin interactions with smoking.\n\nThe brown nodes represent diseases, blue nodes show genes/proteins, and green nodes represent organs (anatomy). Each arrow represents a specific relationship between nodes: “LOCALIZES_DiA” = disease was found to be localized in an anatomy (organ); “EXPRESSES_AuG”, “UPREGULATES_AuG”, “DOWNREGULATES_AuG” mean that the gene is expressed, upregulated, or downregulated in the anatomy (organ), respectively; “INTERACTS_GiG” means that the two genes were found to interact with each other (physically, as proteins); “ASSOCIATES_DaG” means that the gene was found to be associated with the disease; “RESEMBLES_DrD” means that the two diseases were found to occur significantly more often together in MEDLINE articles than would be expected by chance alone. Note that in this setting, the term \"disease\" includes any adverse medical condition, like syndromes, mental disorders, congenital anomalies, and so on.\n\nThe brown nodes represent diseases, blue nodes show genes/proteins, and green nodes represent organs (anatomy). Each arrow represents a specific relationship between nodes: “LOCALIZES_DiA” = disease was found to be localized in an anatomy (organ); “EXPRESSES_AuG”, “UPREGULATES_AuG”, “DOWNREGULATES_AuG” mean that the gene is expressed, upregulated, or downregulated in the anatomy (organ), respectively; “INTERACTS_GiG” means that the two genes were found to interact with each other (physically, as proteins); “ASSOCIATES_DaG” means that the gene was found to be associated with the disease; “RESEMBLES_DrD” means that the two diseases were found to occur significantly more often together in MEDLINE articles than would be expected by chance alone. Note that in this setting, the term \"disease\" includes any adverse medical condition, like syndromes, mental disorders, congenital anomalies, and so on.\n\nThe plot provides information on the recombination rate and linkage disequilibrium between the lead SNP (blue diamond) and other SNPs in the region.\n\nThe plot provides information on the recombination rate and linkage disequilibrium between the lead SNP (blue diamond) and other SNPs in the region.\n\nLeft panel: Setting the minor allele frequency to 0.2 while varying the number of unexposed and exposed triads (unexposed-exposed). Right panel: Setting the number of unexposed and exposed triads to 1100 and 500, respectively, while varying the minor allele frequency. In all analyses, the significance level was 0.05. We varied the maternal RR in exposed triads, so that RRR=RRmat(Exposed). The black curve is the same in both panels because of shared parameters. RRR, relative risk ratio; RRmat, relative risk for an allele inherited from the mother; MAF, minor allele frequency.\n\nQ-Q plots for PoOxSmoke (left) PoOxAlcohol (middle) and PoOxVitamin (right) with 95% pointwise confidence bands. Q-Q, quantile-quantile; PoOxAlcohol, parent-of-origin interactions with alcohol; PoOxSmoke, parent-of-origin interactions with smoking; PoOxVitamin, parent-of-origin interactions with vitamins.\n\nQ-Q plots for PoOxSmoke (left) PoOxAlcohol (middle) and PoOxVitamin (right) with 95% pointwise confidence bands. Q-Q, quantile-quantile; PoOxAlcohol, parent-of-origin interactions with alcohol; PoOxSmoke, parent-of-origin interactions with smoking; PoOxVitamin, parent-of-origin interactions with vitamins.\n\nQ-Q plots for PoOxVitamin with 95% pointwise confidence bands. Q-Q, quantile-quantile; PoOxVitamin, parent-of-origin interactions with vitamins.\n\nAll the top 20 SNPs in the PoOxAlcohol analysis had the same q-value of 0.99 and are therefore not considered here as they are probably false positives (Table 3). All the SNPs in the PoOxSmoke analysis had q-values of around 0.6. Even though these q-values are still quite large, they indicate that around 40% of the SNPs are potentially true PoOxE associations. Among the top 20 SNPs in the PoOxSmoke analysis, two are in the gene for ‘Focadhesin’ (FOCAD), two are in ‘Platelet derived growth factor D’ (PDGFD), two are in ‘Ankyrin 3’ (ANK3), and one is in ‘Fraser syndrome 1’ (FRAS1). Note that associations with PDGFD and ANK3 were also detected in the European analyses (see below). In the PoOxVitamin analysis, only three SNPs had q-values below 0.99, and none of the genes linked to these SNPs have previously been associated with orofacial clefts.\n\nAmong the SNPs with the lowest q-values in the PoOxAlcohol analysis, rs2294035 (q=0.32, p=2.9e-6) and rs4876274 (q=0.76, p=1.3e-5) are in ‘Rho guanine nucleotide exchange factor 10’ (ARHGEF10; GeneCards identifier [GCID]: GC08P001823) (Table 4). The remaining SNPs had q-values above 0.76 and are not considered any further. ARHGEF10 has not previously been linked with orofacial clefts. In the PoOxSmoke analysis, three of the SNPs were in ANK3 (rs3793861: q=0.20, p=2.6e-6; rs7087489: q=0.20, p=3.1e-6; and rs4310561: q=0.67, p=4.0e-5). PoOxE effects in ANK3 were also detected in the analysis of the pooled sample above. To our knowledge, ANK3 has not previously been linked with orofacial clefts, and the same applies to SNP rs10763707 in ‘Lysosome like 1’ (LYZL1; GCID: GC10P029297), which had a q-value of 0.20. In the PoOxVitamin analysis, several of the SNPs shared the same q-value of 0.87 and are not considered any further. The top six SNPs had q-values of 0.44-0.65. Again, none of these genes appear to have any previous connections to clefting. For example, ‘cAMP responsive element binding protein 5’ (CREB5; GCID: GC07P028305) and its network of genes are involved in colorectal cancer42, while ‘R-Spondin 3’ (RSPO3; GCID: GC06P127118) is implicated in tumor development. That said, Park and co-workers reported that RSPO3 acts as an agonist in the canonical Wnt/β-catenin signaling43, a pathway known to be implicated in a wide range of developmental processes, including craniofacial development and homeostasis44–47.\n\nIn the only analysis possible for this ethnic group (PoOxVitamin), all the SNPs had the same q-value of 0.86 (Table 5). They are thus most likely to be false positives and will not be considered any further.\n\nWe chose to focus here on the PoOxE effects detected with SNPs in ANK3 and ARHGEF10. As mentioned above, ANK3 showed up several times among the top PoOxSmoke hits both in the pooled and European analyses, and strong signals for SNPs in ARHGEF10 were detected twice in the European analysis of PoOxAlcohol. We conducted stratified analyses of the effect of the child’s allele, GxE effects, PoO effects, and PoOxE effects for each SNP and haplotype (with haplotypes analyzed both in two-SNP and three-SNP combinations) in these two genes (Table 7). Specifically, we analyzed rs3793861, rs7087489 and rs4310561 in ANK3 that showed PoOxSmoke effects in the European sample, and rs2294035 and rs4876274 in ARHGEF10 that showed PoOxAlcohol effects in the same sample (Table 4). The results showed no child effects or GxSmoke effects for single SNPs in ANK3. By contrast, the p-values were low for all three SNPs in the PoO analyses or PoOxSmoke analyses. This was also the case with the ‘t-a’ allele in the two-SNP combination rs3793861-rs7087489 and the ‘c-t-a’ allele in the three-SNP-combination rs3793861-rs7087489-rs4310561. The other alleles were only associated with PoOxSmoke effects.\n\nFor ARHGEF10, we analyzed the two SNPs that showed PoOxAlcohol effects in the European sample (rs2294035 and rs4876274) but did not discover any effects in either single-SNP or haplotype analyses.\n\nBecause of their low q-values, the genes appearing in the PoOxSmoke and PoOxAlcohol analyses in Table 4 were selected for further analyses using the STRING database, ExpressionAtlas and BGee. However, none of the searches for direct links among the genes yielded any evidence to explain why those genes appeared in our results together.\n\nRegarding the indirect relationships, these are visualized in Figure 1 for relationships between ANK3 and cleft lip (Disease Ontology ID [DOID]: 9296), and, simultaneously, between ANK3 and nicotine dependence (DOID: 0050742). As Hetionet does not include information about smoking, we chose “nicotine dependence” as a proxy. ANK3 is connected to nicotine dependence through several nodes, two of which are particularly noteworthy. First, ANK3 has been reported to be strongly associated with attention-deficit/hyperactivity disorder48, and a connection between attention-deficit/hyperactivity disorder and nicotine dependence has been reported. The connectiom was calculated based on articles listed in MEDLINE, where this pair of conditions co-occured significantly more frequently than would be expected by chance49,50. The second path goes through the gene ‘CRK Like Proto-Oncogene, Adaptor Protein’ (CRKL). It interacts with ANK3 and is downregulated in nicotine dependence. Furthermore, ANK3 is expressed in the telencephalon (the most highly developed part of the forebrain), the embryo, and the head (which are all relevant to CL/P).\n\nFigure 2 shows the indirect relationships between cleft lip, ARHGEF10 and alcohol dependence. Like nicotine dependence in the above analyses, alcohol dependence (DOID: 0050741) is used here as a proxy for maternal alcohol consumption. The only relationships found between ARHGEF10 and cleft lip are the expression of ARHGEF10 in the head and telencephalon. By contrast, there were twelve different relationships between ARHGEF10 and alcohol dependence. However, there were no shared paths connecting cleft lip and alcohol dependence via any of the 12 organs.\n\nThe regional plot for rs3793861 (Figure 3) shows that several SNPs in ANK3 that were not in linkage disequilibrium with rs3793861 had p-values in the range 10-4 to 10-3, which lends support to either ANK3 itself or genes in its vicinity influencing the risk of clefting. However, we did not observe a similar pattern in the regional plot for rs2294035 (Figure 4).\n\nFigure 5 shows that the power does not increase appreciably when the minor allele frequency increases beyond 0.2. However, there is a lot to be gained by increasing the sample size from 500 unexposed and 300 exposed (European, smoke/alcohol) to 1400-600 (pooled, vitamin). Further, the RRRs in the plots are based on changing only the effect of the maternal allele in the exposed triads. The same RRRs could have been achieved in a number of ways, which complicates the interpretation of the RRRs in Table 3–Table 5. Still, if a strong effect is detected with a SNP in a gene, this strengthens the case for its contribution to clefting.\n\n\nDiscussion\n\nThe main aim of this paper was to identify genome-wide PoOxE effects in the larger sample of isolated CL/P, based on the same methodology and GWAS dataset we had previously used in a similar analysis of the smaller sample of isolated CPO27. As with the CPO study, the current analyses benefitted from being based on the largest available GWAS dataset of case-parent triads of orofacial clefts to date. Moreover, data were available for two major ethnicities, European and Asian, which is useful in assessing the generalizability of the findings across different ethnic groups. In the current dataset, however, very few of the Asian mothers reported smoking cigarettes or consuming alcohol during the periconceptional period, thus preventing a comparison of PoOxE effects for these exposures across the two ethnic groups. This is a common impediment to GxE studies, where the number of exposed individuals needs to be large enough for a meaningful analysis51.\n\nRelying on the q-values for assessing the false positive rate, our analyses detected PoOxSmoke effects with SNPs in LYZL1, ANK3, PDGFD, FOCAD and FRAS1, and PoOxAlcohol effects with two SNPs in ARHGEF10. Without formal validation in a comparable and independent replication cohort, it would be premature to dismiss the veracity of the remaining associations merely on the basis of their having too high q-values. High q-values might be the consequence of low statistical power. Likewise, not having previously been linked with orofacial clefts does not necessarily imply that the identified gene is not relevant for clefting. This applies to several genes in our analyses; for example, SHKBP1 and MAPK10 in the PoOxSmoke analysis of the European sample and TJP3 in the PoOxVitamin analysis of the pooled sample. The current study was primed to explore new hypotheses for disease mechanisms and to provide as many of the results as possible so that other researchers with access to similar GWAS datasets would be able to validate the findings presented here. To avoid being overly stringent, we thus presented all the results for the top 20 SNPs in Table 3–Table 5.\n\nDespite an exhaustive literature search, we were unable to find any obvious evidence linking ANK3 and orofacial clefts. The Hetionet results confirmed this lack of a direct connection (Figure 1). ANK3 encodes a member of the Ankyrin family of proteins, whose function is to bind the integral membrane proteins to the spectrin-actin cytoskeleton. This is important for cell motility, activation, proliferation, contact and the maintenance of specialized membrane domains; cellular activities that are also relevant for the proper development of craniofacial structures. For example, Stankewich and colleagues52 showed that the spectrin–ankyrin scaffold is important for cell migration, tissue patterning and organogenesis. Homozygous deletion of the gene encoding αII-spectrin in mice (Spna2) results in craniofacial, neural tube and cardiac anomalies, in addition to retarded intrauterine growth. Figure 3 indicates that several SNPs in ANK3 are potentially associated with clefting.\n\nLike ANK3, ARHGEF10 has not previously been associated with orofacial clefts, and the Hetionet results are consistent with this observation (Figure 2). ARHGEF10 encodes a Rho guanine nucleotide exchange factor that may be involved in neural morphogenesis53. LYZL1 belongs to the family of lysozyme-like proteins that are implicated in sperm function and innate immunity54. According to GeneCards (GCID: GC09P020659), FOCAD encodes a tumor suppressor gene that is highly expressed in the brain. It has also been linked to Alzheimer's disease55. Furthermore, germline deletions in FOCAD are associated with polyposis and colorectal cancer56. Again, as with ANK3 and ARHGEF10 above, there do not seem to be any obvious connections between LYZL1 or FOCAD with clefting.\n\nIn contrast to the above genes, FRAS1 and several members of the platelet-derived growth factor (PDGF) gene family are known to be implicated in orofacial clefts. PDGFD is a member of the PDGF gene family and plays a central role in the PDFG receptor-alpha (PDGFR-α) signaling pathway. More specifically, disruption of Pdgf signaling results in clefting of the palate57. FRAS1 (GCID: GC04P078056) encodes an extracellular matrix protein that plays a critical role in epithelial-mesenchymal interactions during embryonic development58. Loss-of-function mutations in FRAS1 underlie Fraser syndrome, which is characterized by craniofacial, urogenital and respiratory system abnormalities59. Both genes are therefore worthy of further investigations in other isolated orofacial cleft cohorts.\n\nTable 7 shows that in the PoOxSmoke analysis of the two-SNP combination rs3793861-rs7087489 in ANK3, the p-value was slightly lower and the RRR slightly higher than in the corresponding analyses of each individual SNP. A similar pattern was observed in the PoOxAlcohol analyses of the rs2294035-rs4876274 haplotype in ARHGEF10. This indicates that the two-SNP combinations may be driving the effects observed with the individual SNPs.\n\nThe genes PDGFD, CSMD1 and RSUI detected here had previously showed up a study focusing on identifing GxE effects in the same CL/P triads28. In that study, a possible GxVitamin effect was detected with PDGFD and RSUI, and a possible GxAlcohol effect was detected with CSMD1. In the current study, a PoOxSmoke effect was detected with PDGFD and PoOxVitamin effects were detected with RSUI and CSMD1. In other words, only RSUI had the same exposure (vitamin) across the studies. RSUI stands for ‘Ras suppressor protein 1’ and is localized to chromosome 10p13. Its protein product is found at cell–extracellular matrix adhesion sites and has been reported to be involved in supressing v-Ras transformation in the Ras signal transduction pathway60. CSMD1 stands for ‘CUB and Sushi multiple domains 1’ and is localized to chromosome 8p23.2 (GCID: GC08M002953). It is involved in tumor supression, as it has frequently been found to be deleted in many types of cancers61,62. Again, there does not seem to be any obvious connections to clefting.\n\nNone of the top SNPs identified in our previous study focusing on PoOxE effects in CPO triads overlapped with SNPs identified in this study of CL/P27. This is consistent with the assumption that CPO and CL/P are etiologically distinct, so that the lead SNPs may be subtype-specific and differ between the two conditions. However, we detected associations with the ‘cytochrome P450 family 4 subfamily F member 3’ gene (CYP4F3 on chr 19p13.12) in the previous CPO analyses, and with the ‘cytochrome P450 family 46 subfamily A member 1’ gene (CYP46A1 on chr 14q32.2) in the present study. These two genes are members of the cytochrome P450 superfamily of enzymes that are primarily found in liver cells and whose function is to catalyze many reactions involved in the biotransformation of xeno- and endobiotics, and the biosynthesis of cholesterol and lipids, among others63. It is therefore not surprising that these genes would appear in an analysis focusing on smoking, alcohol and vitamin intake.\n\nWe searched Hetionet for indirect links between ANK3 or ARHGEF10 and cleft lip, as well as between ANK3 or ARHGEF10 and nicotine or alcohol dependence, respectively. This approach has several limitations. First, using “nicotine dependence” and “alcohol dependence” in lieu of the actual smoking and alcohol consumption status may introduce some bias. Second, Hetionet is built from a curated set of database information, which means that not all the information, especially the newest, would be available. However, when interpreting our results, we used the source databases to make sure that the connections between the nodes are reliable.\n\nTo conclude, our search for an interaction between a PoO-effect and an environmental exposure for CL/P identified possible relationships between SNPs in ANK3 and maternal smoking, and SNPs in ARHGEF10 and maternal intake of alcohol. There is a possibility that these interactions have a biological basis, although without replication they remain speculative. Our demonstration of the feasibility of identifying complex interactions between relevant environmental exposures and PoO-effects opens new possibilities in the search for the genetic etiology of CL/P.\n\n\nData availability\n\nThe GWAS data are available in the dbGaP database. Additional information regarding the inclusion/exclusion criteria of the study, the ethics statements, data variables, study history, publications, and other documentation related to the study is provided on the dbGaP website.\n\nThe dbGaP database at the National Center for Biotechnology Information, U.S. National Library of Science (NCBI/NLM) can be searched to gain an overview of the cleft dataset used in this study. Entering the dbGaP accession number phs000094.v1.p1 provides access to information regarding the variables, study documents, and datasets. For example, detailed information about the mother’s exposure to alcohol, vitamins, and smoke is provided under the header “Variable Selection”. Information on study questionnaires, institutional review boards and consent forms from each participating cohort can be found under the header “Documents”.\n\nControlled-access data from dbGaP is available only through the dbGaP authorized access portal. There are separate procedures for accessing individual data, depending on whether the researcher is NIH-affiliated (intramural) or not (extramural). In addition, other restrictions apply; e.g., the principal investigator from the applying institution needs to be a permanently employed professor, senior scientist, or equivalent, to submit a data access request. A valid eRA Commons account for logging in to the dbGaP system is also mandatory.\n\n\nSoftware availability\n\n- Source code available from: https://github.com/oeh041/A-genome-wide-scan-of-cleft-lip-triads-identifies-parent-of-origin-interaction-effects-between-ANK3-\n\n- Archived source code at time of publication: https://doi.org/10.5281/zenodo.324131936\n\n- License: CC-BY 4.0", "appendix": "Grant information\n\nThis research was supported by the Bergen Medical Research Foundation [807191], by the Research Council of Norway (RCN) through its Centres of Excellence funding scheme [grant 262700], and by the Biobank Norway II from the RCN [245464].\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe extend our warmest gratitude to the families from so many different countries in the world who opted to contribute to this study. We are also grateful to everybody who contributed to the collection of the data, from recruiters in the field to technicians in the lab.\n\n\nReferences\n\nMosse P, Castilla E: Global registry and database on craniofacial anomalies. In: World Health Organization, Geneva; 2003. Reference Source\n\nBerg E, Haaland ØA, Feragen KB, et al.: Health Status Among Adults Born With an Oral Cleft in Norway. JAMA Pediatr. 2016; 170(11): 1063–1070. PubMed Abstract \| Publisher Full Text\n\nBerg E, Sivertsen Å, Ariansen AM, et al.: Socio-Economic Status and Reproduction among Adults Born with an Oral Cleft: A Population-Based Cohort Study in Norway. PLoS One. 2016; 11(9): e0162196. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStrauss RP: The organization and delivery of craniofacial health Services: the state of the art. Cleft Palate Craniofac J. 1999; 36(3): 189–195. PubMed Abstract \| Publisher Full Text\n\nWehby GL, Cassell CH: The impact of orofacial clefts on quality of life and healthcare use and costs. Oral Dis. 2010; 16(1): 3–10. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRahimov F, Jugessur A, Murray JC: Genetics of nonsyndromic orofacial clefts. Cleft Palate Craniofac J. 2012; 49(1): 73–91. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDixon MJ, Marazita ML, Beaty TH, et al.: Cleft lip and palate: understanding genetic and environmental influences. Nat Rev Genet. 2011; 12(3): 167–178. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJugessur A, Farlie PG, Kilpatrick N: The genetics of isolated orofacial clefts: from genotypes to subphenotypes. Oral Dis. 2009; 15(7): 437–453. PubMed Abstract \| Publisher Full Text\n\nMarazita ML: The evolution of human genetic studies of cleft lip and cleft palate. Annu Rev Genomics Hum Genet. 2012; 13: 263–283. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKousa YA, Schutte BC: Toward an orofacial gene regulatory network. Dev Dyn. 2016; 245(3): 220–232. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBeaty TH, Murray JC, Marazita ML, et al.: Erratum: A genome-wide association study of cleft lip with and without cleft palate identifies risk variants near MAFB and ABCA4. Nat Genet. 2010; 42(8): 727. Publisher Full Text\n\nBirnbaum S, Ludwig KU, Reutter H, et al.: Key susceptibility locus for nonsyndromic cleft lip with or without cleft palate on chromosome 8q24. Nat Genet. 2009; 41(4): 473–477. PubMed Abstract \| Publisher Full Text\n\nCamargo M, Rivera D, Moreno L, et al.: GWAS reveals new recessive loci associated with non-syndromic facial clefting. Eur J Med Genet. 2012; 55(10): 510–514. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGrant SF, Wang K, Zhang H, et al.: A genome-wide association study identifies a locus for nonsyndromic cleft lip with or without cleft palate on 8q24. J Pediatr. 2009; 155(6): 909–913. PubMed Abstract \| Publisher Full Text\n\nMangold E, Ludwig KU, Birnbaum S, et al.: Genome-wide association study identifies two susceptibility loci for nonsyndromic cleft lip with or without cleft palate. Nat Genet. 2010; 42(1): 24–26. PubMed Abstract \| Publisher Full Text\n\nLudwig KU, Mangold E, Herms S, et al.: Genome-wide meta-analyses of nonsyndromic cleft lip with or without cleft palate identify six new risk loci. Nat Genet. 2012; 44(9): 968–971. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGrosen D, Bille C, Pedersen JK, et al.: Recurrence risk for offspring of twins discordant for oral cleft: a population-based cohort study of the Danish 1936-2004 cleft twin cohort. Am J Med Genet A. 2010; 152A(10): 2468–2474. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGrosen D, Chevrier C, Skytthe A, et al.: A cohort study of recurrence patterns among more than 54,000 relatives of oral cleft cases in Denmark: support for the multifactorial threshold model of inheritance. J Med Genet. 2010; 47(3): 162–168. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGrosen D, Bille C, Petersen I, et al.: Risk of oral clefts in twins. Epidemiology. 2011; 22(3): 313–319. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSivertsen A, Wilcox AJ, Skjaerven R, et al.: Familial risk of oral clefts by morphological type and severity: population based cohort study of first degree relatives. BMJ. 2008; 336(7641): 432–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLawson HA, Cheverud JM, Wolf JB: Genomic imprinting and parent-of-origin effects on complex traits. Nat Rev Genet. 2013; 14(9): 609–617. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGjerdevik M, Haaland ØA, Romanowska J, et al.: Parent-of-Origin-Environment Interactions in Case-Parent Triads With or Without Independent Controls. Genet Epidemiol. 2017; 41(7): 664–665.\n\nGuilmatre A, Sharp AJ: Parent of origin effects. Clin Genet. 2012; 81(3): 201–209. PubMed Abstract \| Publisher Full Text\n\nPeters J: The role of genomic imprinting in biology and disease: an expanding view. Nat Rev Genet. 2014; 15(8): 517–530. PubMed Abstract \| Publisher Full Text\n\nThomas D: Gene--environment-wide association studies: emerging approaches. Nat Rev Genet. 2010; 11(4): 259–272. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGjessing HK, Lie RT: Case-parent triads: estimating single- and double-dose effects of fetal and maternal disease gene haplotypes. Ann Hum Genet. 2006; 70(Pt 3): 382–396. PubMed Abstract \| Publisher Full Text\n\nHaaland OA, Jugessur A, Gjerdevik M, et al.: Genome-wide analysis of parent-of-origin interaction effects with environmental exposure (PoOxE): An application to European and Asian cleft palate trios. PLoS One. 2017; 12(9): e0184358. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHaaland ØA, Lie RT, Romanowska J, et al.: A Genome-Wide Search for Gene-Environment Effects in Isolated Cleft Lip with or without Cleft Palate Triads Points to an Interaction between Maternal Periconceptional Vitamin Use and Variants in ESRRG. Front Genet. 2018; 9: 60. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWigginton JE, Cutler DJ, Abecasis GR: A note on exact tests of Hardy-Weinberg equilibrium. Am J Hum Genet. 2005; 76(5): 887–893. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nR Core Team: R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. 2017. Reference Source\n\nSkare O, Jugessur A, Lie RT, et al.: Application of a novel hybrid study design to explore gene-environment interactions in orofacial clefts. Ann Hum Genet. 2012; 76(3): 221–236. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStorey JD, Tibshirani R: Statistical significance for genomewide studies. Proc Natl Acad Sci U S A. 2003; 100(16): 9440–9445. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStorey JD, Bass AJ, Dabney A, et al.: qvalue: Q-value estimation for false discovery rate control. 2015. Reference Source\n\nGjerdevik M, Jugessur A, Haaland ØA, et al.: Haplin power analysis: a software module for power and sample size calculations in genetic association analyses of family triads and unrelated controls. BMC Bioinformatics. 2019; 20(1): 165. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPruim RJ, Welch RP, Sanna S, et al.: LocusZoom: regional visualization of genome-wide association scan results. Bioinformatics. 2010; 26(18): 2336–2337. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHaaland ØA, Romanowska J: A genome-wide scan of cleft lip triads identifies parent-of-origin interaction effects between ANK3 and maternal smoking, and between ARHGEF10 and alcohol consumption. Zenodo. 2019. http://www.doi.org/10.5281/zenodo.3241319\n\nSzklarczyk D, Franceschini A, Wyder S, et al.: STRING v10: protein-protein interaction networks, integrated over the tree of life. Nucleic Acids Res. 2015; 43(Database issue): D447–452. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPetryszak R, Keays M, Tang YA, et al.: Expression Atlas update--an integrated database of gene and protein expression in humans, animals and plants. Nucleic Acids Res. 2016; 44(D1): D746–752. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKomljenovic A, Roux J, Wollbrett J, et al.: BgeeDB, an R package for retrieval of curated expression datasets and for gene list expression localization enrichment tests [version 1; peer review: 1 approved, 1 approved with reservations, 1 not approved]. F1000Res. 2016; 5: 2748. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHimmelstein DS, Lizee A, Hessler C, et al.: Systematic integration of biomedical knowledge prioritizes drugs for repurposing. eLife. 2017; 6: pii: e26726. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBeaty TH, Murray JC, Marazita ML, et al.: A genome-wide association study of cleft lip with and without cleft palate identifies risk variants near MAFB and ABCA4. Nat Genet. 2010; 42(6): 525–529. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nQi L, Ding Y: Involvement of the CREB5 regulatory network in colorectal cancer metastasis. Yi Chuan. 2014; 36(7): 679–684. PubMed Abstract\n\nPark S, Cui J, Yu W, et al.: Differential activities and mechanisms of the four R-spondins in potentiating Wnt/β-catenin signaling. J Biol Chem. 2018; 293(25): 9759–9769. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJin YR, Han XH, Taketo MM, et al.: Wnt9b-dependent FGF signaling is crucial for outgrowth of the nasal and maxillary processes during upper jaw and lip development. Development. 2012; 139(10): 1821–1830. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHe F, Xiong W, Wang Y, et al.: Epithelial Wnt/β-catenin signaling regulates palatal shelf fusion through regulation of Tgfβ3 expression. Dev Biol. 2011; 350(2): 511–519. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHu X, Gao J, Liao Y, et al.: Retinoic acid alters the proliferation and survival of the epithelium and mesenchyme and suppresses Wnt/β-catenin signaling in developing cleft palate. Cell Death Dis. 2013; 4: e898. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJiang Z, Pan L, Chen X, et al.: Wnt6 influences the viability of mouse embryonic palatal mesenchymal cells via the β-catenin pathway. Exp Ther Med. 2017; 14(6): 5339–5344. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCross-Disorder Group of the Psychiatric Genomics Consortium: Identification of risk loci with shared effects on five major psychiatric disorders: a genome-wide analysis. Lancet. 2013; 381(9875): 1371–1379. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHimmelstein DS, Pankov A: Mining knowledge from MEDLINE articles and their indexed MeSH terms. 2015. Publisher Full Text\n\nZhou XZ, Menche J, Barabasi AL, et al.: Human symptoms-disease network. Nat Commun. 2014; 5: 4212. PubMed Abstract \| Publisher Full Text\n\nMcAllister K, Mechanic LE, Amos C, et al.: Current Challenges and New Opportunities for Gene-Environment Interaction Studies of Complex Diseases. Am J Epidemiol. 2017; 186(7): 753–761. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStankewich MC, Cianci CD, Stabach PR, et al.: Cell organization, growth, and neural and cardiac development require αII-spectrin. J Cell Sci. 2011; 124(Pt 23): 3956–3966. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVerhoeven K, De Jonghe P, Van de Putte T, et al.: Slowed conduction and thin myelination of peripheral nerves associated with mutant rho Guanine-nucleotide exchange factor 10. Am J Hum Genet. 2003; 73(4): 926–932. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNarmadha G, Yenugu S: In Silico and Biochemical Characterization of Lysozyme-Like Proteins in the Rat. PLoS One. 2016; 11(9): e0161909. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrockschmidt A, Trost D, Peterziel H, et al.: KIAA1797/FOCAD encodes a novel focal adhesion protein with tumour suppressor function in gliomas. Brain. 2012; 135(Pt 4): 1027–1041. PubMed Abstract \| Publisher Full Text\n\nWeren RD, Venkatachalam R, Cazier JB, et al.: Germline deletions in the tumour suppressor gene FOCAD are associated with polyposis and colorectal cancer development. J Pathol. 2015; 236(2): 155–164. PubMed Abstract \| Publisher Full Text\n\nEberhart JK, He X, Swartz ME, et al.: MicroRNA Mirn140 modulates Pdgf signaling during palatogenesis. Nat Genet. 2008; 40(3): 290–298. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShort K, Wiradjaja F, Smyth I: Let's stick together: the role of the Fras1 and Frem proteins in epidermal adhesion. IUBMB Life. 2007; 59(7): 427–435. PubMed Abstract \| Publisher Full Text\n\nMcGregor L, Makela V, Darling SM, et al.: Fraser syndrome and mouse blebbed phenotype caused by mutations in FRAS1/Fras1 encoding a putative extracellular matrix protein. Nat Genet. 2003; 34(2): 203–208. PubMed Abstract \| Publisher Full Text\n\nCutler ML, Bassin RH, Zanoni L, et al.: Isolation of Rsp-1, a Novel Cdna Capable of Suppressing V-Ras Transformation. Mol Cell Biol. 1992; 12(9): 3750–3756. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMa CQ, Quesnelle KM, Sparano A, et al.: Characterization CSMD1 in a large set of primary lung, head and neck, breast and skin cancer tissues. Cancer Biol Ther. 2009; 8(10): 907–916. PubMed Abstract \| Publisher Full Text\n\nKamal M, Shaaban AM, Zhang L, et al.: Loss of CSMD1 expression is associated with high tumour grade and poor survival in invasive ductal breast carcinoma. Breast Cancer Res Treat. 2010; 121(3): 555–563. PubMed Abstract \| Publisher Full Text\n\nNebert DW, Russell DW: Clinical importance of the cytochromes P450. Lancet. 2002; 360(9340): 1155–1162. PubMed Abstract \| Publisher Full Text" }	[ { "id": "50329", "date": "04 Jul 2019", "name": "Evie Stergiakouli", "expertise": [ "Reviewer Expertise Genetic epidemiology of childhood disorders" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a very interesting paper focusing on identifying parent of origin interaction effects in CL/P using a genome-wide approach. The environmental factors selected are biologically plausible to be involved in cleft and the sample, although modest in size, involved full trios and individuals of non-European ancestry. The authors identify suggestive evidence of interaction for 3 SNPs in ANK3 with smoking and 10 SNPs for ARHGEF10 with alcohol consumption. This study provides proof of principle for their approach and the authors highlight the need for replication in independent samples. The authors follow the results with bioinformatics analysis and the paper is well written and detailed.\n\nMy main concern with the statistical analysis is with the use of directly genotyped SNPs in a genome-wide scan. Can the authors explain why there is no imputation of the directly genotyped SNPs? Imputation would have provided a much larger number of SNPs to work with.\n\nAlso, what is the rational for not using the established genome-wide significance level (510-8)? If anything, the authors should have been more stringent than that since they effectively performed 3 genome-wide analyses (one for each environmental factor).\n\nThe abstract should be made for concise: The 59 individuals of other ancestries were not used for any analyses, so they should not be mentioned in the abstract. The methods section of the abstract does not contain any description of how the performed the analysis and relies heavily on their previous study.\n\nI am not sure if the following statement should be called intriguing: 'Despite this success, the genetic variants identified so far collectively explain only a minor fraction of the total vari- ance attributable to additive genetic effects, which is intriguing considering the more than 70% heritability of CL/P among Europeans17–20'. This is not unusual, and it is the case for many other complex disorder and traits, so I would suggest rephrasing.\n\nThe table titles should include more information to make the paper easier to follow. For example, Table 3 should include that it is a PoOxE analysis.\n\nDiscussion: The authors need to highlight that none of their findings reached genome-wide significance level and given that they have performed more tests that in a standard genome-wide study, the evidence for association is weak. So, their findings require replication and/or meta-analysis in a larger sample. I am not sure I agree with their statement: ‘Without formal validation in a comparable and independent replication cohort, it would be premature to dismiss the veracity of the remaining associations merely on the basis of their having too high q-values.’ All of their findings need to be replicated and should be treated with caution until then.\n\nAnother point that I need to make is that I would have expected the discussion to be more about result interpretation and even trying to suggest possible mechanisms. At the moment the discussion is focused on detailed bioinformatics analysis which although interesting, would be more suited to the results section.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4742", "date": "10 Jul 2019", "name": "Øystein Ariansen Haaland", "role": "Author Response", "response": "We would like to thank Dr. Stergiakouli for her insightful comments, and have addressed her concerns below. We have also made minor changes throughout the manuscript to correct grammatical errors or improve clarity. These are not addressed below, but are highlighted in the new version of the manuscript.Comment:My main concern with the statistical analysis is with the use of directly genotyped SNPs in a genome-wide scan. Can the authors explain why there is no imputation of the directly genotyped SNPs? Imputation would have provided a much larger number of SNPs to work with. Reply:This is a very good observation, and something which we should have addressed in the text.Added to Discussion:“A limitation of this study is that genotypes were not imputed. To avoid Mendelian inconsistencies, the imputation procedure would have had to account for the full triads, as opposed to imputing each sample independently, which has not been done for our data. Instead we conducted post hoc haplotype analyses for combinations of top SNPs located close to each other. This is akin to imputation, in that such an analysis takes into account information about a whole area of DNA, instead of just one SNP.”Comment:Also, what is the rational for not using the established genome-wide significance level (510-8)? If anything, the authors should have been more stringent than that since they effectively performed 3 genome-wide analyses (one for each environmental factor). Reply:In recent years, more and more statisticians have argued for a move away from the term “statistically significant” (e.g., see Amrhein, Greenland & McShane (2019). Retire statistical significance. Nature 567, 305-307 doi: 10.1038/d41586-019-00857-9, and Ronald L. Wasserstein & Nicole A. Lazar (2019) Moving to a World Beyond \"p < 0.05\". Am Stat 73, 1-19 doi: 10.1080/00031305.2019.1583913), which is why we have avoided using a fixed p-value as threshold for significance. Therefore, we focused on the q-value. However, for transparency we still report p-values, as well as effect estimates with 95% confidence intervals.Added to Methods:“In accordance with recent recommendations by the American Statistical Association and others 34,35, we did not consider a fixed p-value as a threshold for statistical significance.” Comment: The abstract should be made for concise: The 59 individuals of other ancestries were not used for any analyses, so they should not be mentioned in the abstract. Reply:We mention these 59 individuals because they were part of the pooled analyses, although they were not analyzed as a group. This is perhaps not clear from the text. In response, we have added a sentence to clarify this in the text.Added to Methods (in the paper, not the abstract):“The 59 families with other ethnicities were not analyzed as a group due to the small sample size, but they were included in the pooled analysis.” Comment:The methods section of the abstract does not contain any description of how the performed the analysis and relies heavily on their previous study.Reply:Thank you for pointing this out. Text changed in Abstract:“We recently developed new methods to identify parent-of-origin (PoO) interactions with environmental exposures (PoOxE) and now apply them to data from a genome-wide association study (GWAS) of families with children born with isolated CL/P.”“The methodology applied in the analyses detects if PoO effects are different across environmental strata, and is implemented in the R-package Haplin.” Comment:I am not sure if the following statement should be called intriguing: 'Despite this success, the genetic variants identified so far collectively explain only a minor fraction of the total variance attributable to additive genetic effects, which is intriguing considering the more than 70% heritability of CL/P among Europeans17–20'. This is not unusual, and it is the case for many other complex disorder and traits, so I would suggest rephrasing. Reply:Thank you for pointing this out.Text changed in Introduction:“Despite this success, the genetic variants identified so far collectively explain only a minor fraction of the total variance attributable to additive genetic effects, even though the heritability of CL/P among Europeans is more than 70% 17-20.” Comment:The table titles should include more information to make the paper easier to follow. For example, Table 3 should include that it is a PoOxE analysis. Reply:Thank you for this comment.Title changed in Table 3:“The top 20 single-nucleotide polymorphisms (SNPs) sorted by p-value in the pooled PoOxE analysis.”Title changed in Table 4:“The top 20 single-nucleotide polymorphisms (SNPs) sorted by p-value in the European PoOxE analysis.”Title changed in Table 7:“Stratified analyses of the top single-nucleotide polymorphisms (SNPs) and haplotypes in ANK3 (PoOxSmoke) and ARHGEF10 (PoOxAlcohol).” Comment:Discussion: The authors need to highlight that none of their findings reached genome-wide significance level and given that they have performed more tests that in a standard genome-wide study, the evidence for association is weak. So, their findings require replication and/or meta-analysis in a larger sample. I am not sure I agree with their statement: ‘Without formal validation in a comparable and independent replication cohort, it would be premature to dismiss the veracity of the remaining associations merely on the basis of their having too high q-values.’ All of their findings need to be replicated and should be treated with caution until then. Reply:Thank you for pointing this out.Text changed in Discussion:“Relying on the q-values for assessing the false positive rate, our analyses detected possible PoOxSmoke effects with SNPs in LYZL1, ANK3, PDGFD, FOCAD and FRAS1, and possible PoOxAlcohol effects with two SNPs in ARHGEF10. Without formal validation in a comparable and independent replication cohort, it would be premature to accept these associations as true PoOxE effects.” Comment:Another point that I need to make is that I would have expected the discussion to be more about result interpretation and even trying to suggest possible mechanisms. At the moment the discussion is focused on detailed bioinformatics analysis which although interesting, would be more suited to the results section.Reply:Thank you for this suggestion.Added to Discussion:“A possible mechanism for a PoO effect is genomic imprinting 21. This occurs when DNA methylation in the germline causes the expression of alleles to be silenced depending on their parental origin. Maternal environmental exposures that affect methylation patterns may also affect paternally and maternally inherited alleles differently. Furthermore, a PoO effect that is not affected by an environmental exposure will not be detected in our PoOxE analysis. Hence, a detected PoOxE effect may have a better chance than a PoO or GxE effect of uncovering a true causal relationship involving genomic imprinting.”" } ] }, { "id": "50328", "date": "25 Jul 2019", "name": "Yongchu Pan", "expertise": [ "Reviewer Expertise Genetics of craniofacial diseases" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors conducted a genome-wide interaction analysis with smoking, alcohol, and vitamin on cleft lip triads. The analysis was based on 1,908 nuclear families from dbGaP database with both genotyping and environmental exposure status. Their results suggested that three SNPs in ANK3 has a PoOxSmoke effect and two SNPs in ARHGEF10 has a PoOxAlcohol effect. The results may provide evidence for identifying interactions between relevant environmental exposures and PoO effects. Totally, this paper was very well written; however, there are some concerns that should be addressed to improve the paper as detailed below:\nThe GWAS data are available in the dbGaP database (the dbGaP accession number: phs000094.v1.p1). Since the imputed data were provided from dbGaP, it is more appropriate to use that data for interaction analysis to find more meaningful results in this study.\n\nIt seems the author only presented the process on quality control for SNPs, and the process on quality control for individuals should be added in the part of Methods.\n\nIn the present study, the author focused the q-value rather than p-value, which is more widely accepted as known. It is confused that the cut off of q-value chosen in each interaction analysis on different population is unclear.\n\nIt is unclear that the cutoff of q-value in each interaction analysis on different population. The author needs to determine a clear threshold for screening, which should be more strict.\n\nThere are a lot of tables in the article. It is recommended to add the forest plots of the interaction analysis in order to make the results more intuitive.\n\nI am not sure if it is appropriate to state ‘After quality control, 341,191 SNPs remained from the original 569,244.’ in Methods section from abstract, because that seems to be part of the result.\n\nAll the population and analysis were based on nuclear families with cleft lip with/without cleft palate, it is proper to replace ‘cleft lip’ with ‘cleft lip with/without cleft palate’ in the title of this paper.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-960
https://f1000research.com/articles/7-1987/v1	30 Dec 18	{ "type": "Research Note", "title": "Research on Babesia: A bibliometric assessment of a neglected tick-borne parasite", "authors": [ "Alfonso J. Rodriguez-Morales", "D. Katterine Bonilla-Aldana", "Juan Pablo Escalera-Antezana", "Lucia Elena Alvarado-Arnez", "D. Katterine Bonilla-Aldana", "Juan Pablo Escalera-Antezana", "Lucia Elena Alvarado-Arnez" ], "abstract": "Given the emergence and reemergence of tick-borne diseases, here we assessed the publishing patterns of research focused on Babesia. We also discuss the implications for the articles published in the last decade, and how more clinical and epidemiological information concerning Babesia is still required. The findings of this article would be useful to define research priorities about Babesia and diagnose the important of scientific production on this pathogen.", "keywords": [ "Babesia", "tick-borne disease", "epidemiology", "public health", "bibliometric" ], "content": "Introduction\n\nBabesiosis is a zoonotic disease with a global distribution; it is mainly transmitted by ticks of different genera (e.g. Rhipicephalus spp. and Dermacentor spp.) and diverse species1. It is caused by infection of the erythrocytes of mammals by Babesia species, which are Apicomplexa protozoa of the suborder Piroplasmidea and the family Babesiidae2. The vector role of ticks for these parasites was discovered by Smith and Kilbourne in 1893, who were the first to demonstrate its transmission3. The first case was described by Skaraballo and occurred in 1957 in Zagreb, Croatia4.\n\nHuman babesiosis is not under surveillance and notification in most countries, including those with autochthonous incidence vector-borne diseases. However, studies show that their vectors are widely distributed in tropical and subtropical areas3. Research is fundamental to better comprehending this disease. The relevance of bibliometric evaluations on emerging and reemerging disease has been previously described5–7 as they can contribute in the understanding on how the global scientific and health communities respond to outbreaks8. Herein, our objective was to use bibliometric approaches to analyze Babesia research.\n\n\nMethods\n\nA bibliometric evaluation was performed focusing on Babesia scientific bibliography. Six main databases were used for retrieving information: Science Citation Index Expanded (SCI-E), Scopus, Medline, LILACS, SciELO and Google Scholar.\n\nFor the search pipeline we used the following combination of key words (MeSH, Medical Subject Headings): “Babesia” AND “Latin America”, “Babesia” AND “Argentina”, “Babesia” AND “Colombia”, and this strategy was maintained including the name of each country as a keyword. Also, “Babesiosis” was used as a substitute of Babesia to increase the number of results. Regarding the type of publications, we decided to include original papers, review articles, case reports and editorials, which were further stratified according to publication year and the name and institution to which the main author was affiliated at the time of publishing. This analysis included results obtained up to December 1, 2018.\n\nData summaries for quantitative variables (number of articles, articles per country, articles per year or periods, citations and H index) were expressed as means and interquartile ranges (IQRs), and for qualitative variables proportions are reported.\n\n\nResults\n\nOverall, 78,137 Babesia-associated items resulted from the initial screening of publications. From Google Scholar 62,100 articles (25% USA, 24.9% South Africa, 18.5% Japan) were recovered, followed by Scopus, with 6,272 articles (25.4% from USA, 8.5% Japan, 7.2% UK), and Medline with 5.045 articles (13.7% USA, 10.1% Japan and 5.2% China) (Table 1). From Web of Science, 4,330 publications were retrieved (28.06% from USA, 11.4% Japan and 7.37% Brazil), followed by LILACS with 202 articles (29.2% Brazil, 2.4% Mexico, 1.9% USA) and SciELO with 188 articles (26.6% Brazil, 3.1% Mexico) (Table 1). Considering the Medline database, the number of research articles on Babesia increased above 100 publications per year only after 2004 (Figure 1).\n\nIn the case of Scopus, the median number of articles published each year as of 1970 was only one (IQR: 0-3), from 1970 until 1995 this number increased to 64 (IQR: 56-73) and from 1996 till 2018 was 188 (IQR: 115–271) (Figure 2). At Scopus 134 countries contributed a minimum of one paper over the study period. For SCI-E, the annual median number of articles reported from 1996 until 2018 was of 99 (IQR: 96-103) (Figure 3), with at least one article published from 129 countries during the study period.\n\n“Obihiro University” in Hokkaido, Japan, was the institution with the most productive research in Scopus, and “Igarashi, I” was the author with the largest record in Babesia research, with 210 articles (Figure 4 and Figure 5). At Web of Sciences, the H index for the topic is 88, with 70,950 citations, reaching 7,734 citations in 2017 (Figure 6).\n\nThe raw data generated in this study are available on OSF9.\n\n\nDiscussion\n\nThe results presented here show that the USA and Japan have primary roles in Babesia research, with USA leading the scientific production with nearly quarter of the published articles, followed by Japan and the UK (Table 1). Certainly, in USA, tickborne disease occurrence is frequent especially in certain areas and months over the year. Tickborne diseases such as babesiosis are commonly reported in Northeastern states as well in the upper Midwest, often with higher incidence in summer. In addition, blood transfusions is still a matter of concern, even in the USA10–13. In countries in Asia, such as Japan, human babesiosis was not reported until fairly recently (1999), when a symptomatic case was describe in Kobe City, Hyogo Prefecture, Japan14,15; however, since then research has significantly increased in this country. Authors from UK have collaborated with research with others from endemic countries. However, in 2006 and 2016, two cases of autochthonous canine babesiosis were reported in the UK. Since November 2015, there have been at least three more cases of canine babesiosis in untraveled dogs from Essex, all were confirmed B. canis infections by PCR. Dermacentor reticulatus ticks were found on the dogs16.\n\nOne of the relevant aspects surrounding babesiosis is that there are not yet licensed human prophylactic vaccines, and treatment alternative remain limited. Two commonly used antimicrobial regimes are highly effective: the combination of atovaquone and azithromycin and the combination of clindamycin and quinine17. Thus, most preventive measures are needed to reduce the risk of infection from ticks and wild and domestic reservoirs (e.g. rats).\n\nBibliometric analyses contribute an objective vision of the scientific activity of a country or a region, in an investigative area. In the particular case of infectious diseases, there are different reports about its utility5–8, especially in emerging infectious diseases18–20, being possible to establish and to compare the amount of scientific production in journals, institutions, and authors publishing about a certain issue; this would allow establishment of a plan in terms of scientific policy as well in other matters21. No previous bibliometric studies about babesiosis or Babesia have been found in the consulted bibliographical scientific databases.\n\nIn conclusion, it is time to translate research findings into effective control of babesiosis. As occurs with other emerging diseases, research leading to vaccinal or effective therapeutic options are of utmost importance. Tick-borne pathogens such as Babesia and others with even clearer epidemic potential need to be researched more and to be prioritized with effective interventions to reduce their negative impact.\n\n\nData availability\n\nRaw bibliometric data generated in this study are available on OSF. DOI: https://doi.org/10.17605/OSF.IO/ER9UP9.", "appendix": "Grant information\n\nThe author(s) declared that not grants were involved in supporting this work.\n\n\nReferences\n\nRodriguez-Morales AJ: Epidemiología de la Babesiosis: Zoonosis emergente. Acta Cient Estud. 2007; 5(4): 132–8. Reference Source\n\nBeugnet F, Moreau Y: Babesiosis. Rev Sci Tech. 2015; 34(2): 627–39. PubMed Abstract\n\nGray JS, Estrada-Peña A, Zintl A: Vectors of Babesiosis. Annu Rev Entomol. 2018. PubMed Abstract \| Publisher Full Text\n\nBrasseur P, Gorenflot A: Human babesial infections in Europe. Rocz Akad Med Bialymst. 1996; 41(1): 117–22. PubMed Abstract\n\nAlmeida-Guerrero A, Olaya-Gómez JC, Sánchez-Ramírez N, et al.: Mitigation of the global impact of Lassa fever: Have we investigated enough about this Arenavirus? - A bibliometric analysis of Lassa Fever research. Travel Med Infect Dis. 2018; 24: 13–4. PubMed Abstract \| Publisher Full Text\n\nBallabeni A, Boggio A: Publications in PubMed on Ebola and the 2014 outbreak. F1000Res. 2015; 4: 68. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRodríguez-Morales AJ, Ramírez-Jaramillo V, Patiño-Barbosa AM, et al.: Severe fever with thrombocytopenia syndrome - A bibliometric analysis of an emerging priority disease. Travel Med Infect Dis. 2018; 23: 97–8. PubMed Abstract \| Publisher Full Text\n\nCulquichicón C, Cardona-Ospina JA, Patiño-Barbosa AM, et al.: Bibliometric analysis of Oropouche research: impact on the surveillance of emerging arboviruses in Latin America. F1000Res. 2017; 6: 194. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRodriguez-Morales AJ: Bibliometric Data of Babesia. OSF. 2018. http://www.doi.org/10.17605/OSF.IO/ER9UP\n\nHajicharalambous C, Rattu M, Leuchten S: A Walk in the Park: A Case of Babesiosis in the South Bronx. Clin Pract Cases Emerg Med. 2018; 2(1): 61–3. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJajosky RP, Jajosky AN: Is babesiosis the most common transfusion transmitted infection in the United States of America? The answer is not simple! Transfus Apher Sci. 2017; 56(4): 609–10. PubMed Abstract \| Publisher Full Text\n\nLinden JV, Prusinski MA, Crowder LA, et al.: Transfusion-transmitted and community-acquired babesiosis in New York, 2004 to 2015. Transfusion. 2018; 58(3): 660–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStein E, Elbadawi LI, Kazmierczak J, et al.: Babesiosis Surveillance - Wisconsin, 2001-2015. MMWR Morb Mortal Wkly Rep. 2017; 66(26): 687–91. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMatsui T, Inoue R, Kajimoto K, et al.: [First documentation of transfusion-associated babesiosis in Japan]. Rinsho Ketsueki. 2000; 41(8): 628–34. PubMed Abstract\n\nTsuji M, Wei Q, Zamoto A, et al.: Human babesiosis in Japan: epizootiologic survey of rodent reservoir and isolation of new type of Babesia microti-like parasite. J Clin Microbiol. 2001; 39(12): 4316–22. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCook S, English K, Humm K: Autochthonous babesiosis in the United Kingdom. J Small Anim Pract. 2016; 57(6): 332. PubMed Abstract \| Publisher Full Text\n\nVannier E, Gewurz BE, Krause PJ: Human babesiosis. Infect Dis Clin North Am. 2008; 22(3): 469–88 viii-ix. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRodriguez-Morales AJ, Ramirez-Jaramillo V, Sánchez-Carmona D, et al.: Kyasanur forest disease: Another flavivirus requiring more research? Results of a bibliometric assessment. Travel Med Infect Dis. 2017; 19: 68–70. PubMed Abstract \| Publisher Full Text\n\nRodriguez-Morales AJ, Sabogal-Roman JA, Alvarez-Moreno CA: What has been researched about MDR-Candida auris? A bibliometric analysis on the 'new kid on the block' in hospital-associated infections. J Infect Public Health. 2018; 11(2): 295–6. PubMed Abstract \| Publisher Full Text\n\nVera-Polania F, Muñoz-Urbano M, Bañol-Giraldo AM, et al.: Bibliometric assessment of scientific production of literature on chikungunya. J Infect Public Health. 2015; 8(4): 386–8. PubMed Abstract \| Publisher Full Text\n\nVera-Polania F, Perilla-Gonzalez Y, Martinez-Pulgarin DF, et al.: Bibliometric assessment of the Latin-American contributions in dengue. Recent Pat Antiinfect Drug Discov. 2014; 9(3): 195–201. PubMed Abstract \| Publisher Full Text" }	[ { "id": "42304", "date": "08 Feb 2019", "name": "Stalin Vilcarromero", "expertise": [ "Reviewer Expertise Clinical and epídemiological research in Vector Borne Disease" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn the manuscript entitled “Research on Babesia: A bibliometric assessment of a neglected tick-borne parasite” the authors try to evaluate the previous bibliometric research regarding babesiosis in the world. It has recognized the value of this type of study because it helps to identify the importance of a country, institution or researcher in solving problems based on scientific evidence. Most of them describe bibliometrics in their papers considering bibliometric variables such as number of citations, author participation in research production, author and co-authorship analysis with VOSviewer1, the H-index, geographical distribution of that disease by countries, the amount and intensity of their international collaboration, analysis of that research based on the frequency of the words used in the title of the articles2, number of publications with intra-country collaboration, number of publications with inter-country collaboration3,4 etc, and usually, the literature was retrieved using only one database (Scopus, Medline, etc) which may give the advantage to let them analyze this in depth.\nHowever, when the number of publications is normalized by population1, by gross domestic product, and by gross national income per capita5, health expenditure6, scientific collaboration7 or other important variable (epidemiology variables such as prevalence, incidence8, endemic versus non-endemic, etc) it makes more relevant the study. On the other hand, it is known the capacity of databases such Scopus, Medline, Web of Science and Scielo, and the authors may decide to use only one giving more details to the analysis.\nIn this case, considering that Babesiosis is a neglected disease, of importance in several countries as it has been described in the manuscript, so this topic deserve still more research, so I consider that this bibliometric analysis would be important for the scientific community. However, it would also be important to normalize the number of publications (including some of those variables mentioned above), include other bibliometric variables such as H-index.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4726", "date": "18 Jul 2019", "name": "Alfonso Rodriguez-Morales", "role": "Author Response", "response": "Dear Dr. Vilcarromero Thanks for your valuable comments. Regards them, we would like to comment and discuss, in the context of the submission a new revised version (version 2). In the manuscript entitled “Research on Babesia: A bibliometric assessment of a neglected tick-borne parasite” the authors try to evaluate the previous bibliometric research regarding babesiosis in the world. It has recognized the value of this type of study because it helps to identify the importance of a country, institution or researcher in solving problems based on scientific evidence. Most of them describe bibliometrics in their papers considering bibliometric variables such as number of citations, author participation in research production, author and co-authorship analysis with VOSviewer1, the H-index, geographical distribution of that disease by countries, the amount and intensity of their international collaboration, analysis of that research based on the frequency of the words used in the title of the articles2, number of publications with intra-country collaboration, number of publications with inter-country collaboration3,4 etc, and usually, the literature was retrieved using only one database (Scopus, Medline, etc) which may give the advantage to let them analyze this in depth. We performed a bibliometric study; we did not evaluate previous bibliometric research. As until the present date, there are no records of previous published bibliometric study in this subject. Regarding the “method”, there is still no consensus on reporting or performing bibliometric studies. Some studies analyze only one database, but this would lead to certain bias, as e.g. WoS have a limited geographical coverage, compared to Scopus. Or Medline is limited in number of journals of certain biomedical areas. The advantage of having different databases is to have a wider coverage and avoid geographical bias, as we intended to do. Secondly, we have published a Research Note (a brief article), not an Original Article. Then extension is quite different. In reference to variables, we used the most common ones utilized in bibliometric studies. However, when the number of publications is normalized by population1, by gross domestic product, and by gross national income per capita5, health expenditure6, scientific collaboration7 or other important variable (epidemiology variables such as prevalence, incidence8, endemic versus non-endemic, etc) it makes more relevant the study. On the other hand, it is known the capacity of databases such Scopus, Medline, Web of Science and Scielo, and the authors may decide to use only one giving more details to the analysis. Although the proposed relationships would be very relevant, this is a research note for a descriptive bibliometric study. Such analysis would result in an extended analytical bibliometric study for a Full-Length Original Article. Even more, some of the suggested variables are not standardized for many countries nor available for the whole period of years and even more publicly available. For example, as we clearly stated, babesiosis is not under surveillance in most countries. Then, this makes highly unlikely to have epidemiological indicators to make such correlations. However, we included in the Table 1 an adjustment per capita of the number of publications per million population of the countries. In the case of USA, where data from 2011 to 2015 was available, we analyzed and compared the number of reported cases in that period with the number of publications at Scopus, WoS and PubMed during the same time. That data from USA, was published this year (May 2019), and was now cited by us, but not previously available, when we performed the bibliometric study. For more years, and other countries, this is not possible. In addition, we have access to data from Wisconsin, USA, which was included and analyzed (Figure 12). In this case, considering that Babesiosis is a neglected disease, of importance in several countries as it has been described in the manuscript, so this topic deserve still more research, so I consider that this bibliometric analysis would be important for the scientific community. However, it would also be important to normalize the number of publications (including some of those variables mentioned above), include other bibliometric variables such as H-index. We are grateful to the reviewer in the appreciation that the bibliometric analysis is of relevance for the scientific community. In reference of the mentioned variables, the H index is already included in the manuscript (third paragraph of the results description)." } ] }, { "id": "44953", "date": "26 Feb 2019", "name": "Jeremy S. Gray", "expertise": [ "Reviewer Expertise Parasitologist", "with special interest in babesiosis and tick-borne diseases" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article attempts to assess the bibliographic status of Babesia parasites with the declared objective of identifying research priorities in order to achieve effective prevention and control of babesiosis. The authors have produced publication data from various sources showing trends over the years and also by citation, author, institute and country.\nUnfortunately the article has major deficiencies. The most obvious of these is that the human and animal versions of the disease have been conflated so that the data are more or less meaningless. The economic impact, research priorities and research constraints are very different in veterinary compared with medical babesiosis. In fact it is possible to argue that even the parasites are different, since the vast majority of human cases are caused by a parasite (Babesia microti) that is only distantly related to those prevalent in veterinary babesiosis (Babesia sensu stricto), and there are differences in their biology such as presence or absence of transovarial transmission, sensitivity to antibabesials, availability of in vitro cultures etc.\nAdditionally, it is difficult to see the point of counting numbers of articles published by different countries, authors, institutions etc. Cross comparison of these data is invidious since different interests and time scales are involved. Such data may tell the reader something about where there has been sufficient interest for research funding but nothing about the nature of the research, which is necessary to identify areas of neglect. It would have been much more useful to break the data down by topic. For example, how many articles are in the area of pure immunological research, usually involving rodents, how many address therapeutic issues, how many vaccination, how many epidemiology etc. Only then would it be possible to see where the gaps are, particularly in relation to practical measures, particularly if accompanied by analytical comments. The superficial approach of this article certainly does not, especially when there has been no attempt to explain the trends presented in the figures.\nThe authors have identified some research areas that require more attention, for example blood transfusion infection in the USA, development of vaccines (presumably molecular), development of new antibabesials, but have not provided the necessary context or evidence for these conclusions.\nSome important issues that the authors seem to have ignored completely, include the development and successful use of live vaccines for cattle babesiosis over a long period of time, the prodigious, but failed efforts, to produce molecular vaccines against cattle babesiosis (which indicates the very great difficulty involved in the development of vaccines for human use), the change in direction and emphasis of babesiosis research in general with the discovery of B. microti in the US, and epidemiological issues such as the extension of the geographical range of infections, although briefly touched on in relation to the UK, for reasons that are not clear.\nThe references make little mention of established authorities in the topic and there are at least five instances of self-citation. Finally there are many examples of poor sentence construction (e.g. the last part of the last sentence in the abstract, inaccurate statements (e.g. the first sentence in the Introduction and the first sentence in the second paragraph), unnecessary sentences (e.g. the third sentence in the second paragraph) etc. There are more of all these in the Discussion.\nOverall, the impression gained is that the authors have made use of readily available metrics on the internet, to present data that appear to have no useful meaning and have not attempted to analyze the data to achieve their stated objectives.\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Not applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [ { "c_id": "4727", "date": "18 Jul 2019", "name": "Alfonso Rodriguez-Morales", "role": "Author Response", "response": "Dear Dr. Gray Thanks for your comments. We have revised and improved the manuscript. In regard to your observations, we would like to comment and discuss, in the context of the submission a new revised version (version 2). This article attempts to assess the bibliographic status of Babesia parasites with the declared objective of identifying research priorities in order to achieve effective prevention and control of babesiosis. The authors have produced publication data from various sources showing trends over the years and also by citation, author, institute and country. The objective of this bibliometric study was not to “identifying research priorities in order to achieve effective prevention and control of babesiosis”, which would be impossible from this type of study. What we really declared is “The findings of this article would be useful to define research priorities about Babesia and diagnose the important of scientific production on this pathogen” (final line of the abstract) Unfortunately the article has major deficiencies. The most obvious of these is that the human and animal versions of the disease have been conflated so that the data are more or less meaningless. The economic impact, research priorities and research constraints are very different in veterinary compared with medical babesiosis. In fact it is possible to argue that even the parasites are different, since the vast majority of human cases are caused by a parasite (Babesia microti) that is only distantly related to those prevalent in veterinary babesiosis (Babesia sensu stricto), and there are differences in their biology such as presence or absence of transovarial transmission, sensitivity to antibabesials, availability of in vitro cultures etc. We thank the reviewer for the observation, although we consider that today, the vision of zoonoses should be an integrated one. Then, having separated human and animal babesiosis, to us, is not rationale. Babesiosis is one zoonotic disease, no matter the host. The work on babesiosis should be together between veterinarians and human physicians, working in the interphase that zoonoses, such as babesiosis, provide. One World, One Health. Additionally, it is difficult to see the point of counting numbers of articles published by different countries, authors, institutions etc. Cross comparison of these data is invidious since different interests and time scales are involved. Such data may tell the reader something about where there has been sufficient interest for research funding but nothing about the nature of the research, which is necessary to identify areas of neglect. It would have been much more useful to break the data down by topic. For example, how many articles are in the area of pure immunological research, usually involving rodents, how many address therapeutic issues, how many vaccination, how many epidemiology etc. Only then would it be possible to see where the gaps are, particularly in relation to practical measures, particularly if accompanied by analytical comments. The superficial approach of this article certainly does not, especially when there has been no attempt to explain the trends presented in the figures. Bibliometric studies quantify and compare the scientific output on specific and general topics. That is the idea and the design, that has been a base consensus in this type of studies. Nevertheless, we have included additional information in regards the areas of research related to Babesia as well as the funding sponsors, from both Scopus and Web of Sciences. The authors have identified some research areas that require more attention, for example blood transfusion infection in the USA, development of vaccines (presumably molecular), development of new antibabesials, but have not provided the necessary context or evidence for these conclusions. That would be more related to other kind of studies. If required, a specific study (e.g. review) about that should be designed, to address those specific topics. Some important issues that the authors seem to have ignored completely, include the development and successful use of live vaccines for cattle babesiosis over a long period of time, the prodigious, but failed efforts, to produce molecular vaccines against cattle babesiosis (which indicates the very great difficulty involved in the development of vaccines for human use), the change in direction and emphasis of babesiosis research in general with the discovery of B. microti in the US, and epidemiological issues such as the extension of the geographical range of infections, although briefly touched on in relation to the UK, for reasons that are not clear. We agree with the reviewer that such points would be of interest. But our study characterizes as a bibliometric research note, and not a systematic review, or a bibliometric study about vaccines. Therefore, that was not an objective of our study. The references make little mention of established authorities in the topic and there are at least five instances of self-citation. Finally there are many examples of poor sentence construction (e.g. the last part of the last sentence in the abstract, inaccurate statements (e.g. the first sentence in the Introduction and the first sentence in the second paragraph), unnecessary sentences (e.g. the third sentence in the second paragraph) etc. There are more of all these in the Discussion. Thank you very much for the observation. In this revised version, we attended several of such issues in order to improve the manuscript. Overall, the impression gained is that the authors have made use of readily available metrics on the internet, to present data that appear to have no useful meaning and have not attempted to analyze the data to achieve their stated objectives. As previously indicated, in this revised version the manuscript has been improved. Although we would like to clarify that: i) the metrics presented in this research note were not available on the internet and ii) the databases that were evaluated are significant sources for bibliometric studies, to whom (Scopus and SCI-E Web of Knowledge) the Universidad Tecnológica de Pereira, in Colombia has valid subscriptions." } ] }, { "id": "43816", "date": "28 Feb 2019", "name": "Cristina Casalone", "expertise": [ "Reviewer Expertise Diagnostics on vector borne disease", "neuropathology", "surveillance program" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript submitted by Rodríguez Morales et al. represents a bibliometric evaluation on Babesia, in order to contribute to understanding this neglected zoonosis and addressing future research and control strategies. Bibliometric evaluation is an excellent tool to obtain objective information about specific areas of research and support the adoption of strategic decisions. In detail, this study provides and summaries data on the research activity on Babesia worldwide. It shows that the main country involved in the research on Babesia is USA, where babesiosis is a notifiable disease since 2011 (CDC) and most human cases have been reported. Of interest the strong research activity of institutions and countries as Japan and UK, in which Babesia represents a new emerging problem both in animals and humans. This study highlights the increased research activity on this neglected zoonosis, considered of growing importance in several countries and the need of further studies addressed to preventive and therapeutic aspects. The manuscript, proposed as a research note, is well structured, the statistical analysis and its interpretation is sufficient, results and discussion appropriate. To fulfil F1000Research criteria (“Research note … can be reported with one or two illustrations (figures/tables)”), authors could reduce the number of figures/graphs.\nI suggest the following minor revisions in the text:\nIn the Introduction:\nI suggest to add Ixodes spp as tick genera involved in the transmission of Babesia to humans. Indeed in USA most reported human cases are attributed to B. microti transmitted to people by Ixodes scapularis. Moreover, most European human cases are caused by B. divergens and B. venatorum, primarly transmitted by Ixodes ricinus1. The authors should specify that the first case described in Croatia in 1957 by Skaraballo, refers to a “human” case. Moreover I suggest adding a sentence regarding the role of animal reservoirs and their distribution that contributes (as the presence of vectors) in the maintenance of the transmission cycle.\n\nIn the Methods: You could clarify which and/or how many countries have been used as keyword for the search pipeline\nFigure 6: In the caption: the citation trends is from Web of Science (as reported in the results), not Scopus.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4728", "date": "18 Jul 2019", "name": "Alfonso Rodriguez-Morales", "role": "Author Response", "response": "Dear Dr. Casalone Thanks for your valuable comments. Regards them, we would like to comment and discuss, in the context of the submission a new revised version (version 2). The manuscript submitted by Rodríguez Morales et al. represents a bibliometric evaluation on Babesia, in order to contribute to understanding this neglected zoonosis and addressing future research and control strategies. Bibliometric evaluation is an excellent tool to obtain objective information about specific areas of research and support the adoption of strategic decisions. In detail, this study provides and summaries data on the research activity on Babesia worldwide. It shows that the main country involved in the research on Babesia is USA, where babesiosis is a notifiable disease since 2011 (CDC) and most human cases have been reported. Of interest the strong research activity of institutions and countries as Japan and UK, in which Babesia represents a new emerging problem both in animals and humans. This study highlights the increased research activity on this neglected zoonosis, considered of growing importance in several countries and the need of further studies addressed to preventive and therapeutic aspects. We are thankful for your comments. In this revised version, we have included more information such as, an adjustment per capita of the number of publications per million population of the countries (Table 1). In the case of USA, where data from 2011 to 2015 was available, we analyzed and compared the number of reported cases in that period with the number of publications at Scopus, WoS and PubMed during the same time. The manuscript, proposed as a research note, is well structured, the statistical analysis and its interpretation is sufficient, results and discussion appropriate. To fulfil F1000Research criteria (“Research note … can be reported with one or two illustrations (figures/tables)”), authors could reduce the number of figures/graphs. Thank you for the comments. Per request from the other reviewers, we actually have extended on analyses, data and figures in order to respond to their inquiries. I suggest the following minor revisions in the text:In the Introduction:I suggest to add Ixodes spp as tick genera involved in the transmission of Babesia to humans. Done. Included (Second line in Introduction section).Indeed, in USA most reported human cases are attributed to B. microti transmitted to people by Ixodes scapularis. Moreover, most European human cases are caused by B. divergens and B. venatorum, primarly transmitted by Ixodes ricinus1.Done. Included. The authors should specify that the first case described in Croatia in 1957 by Skaraballo, refers to a “human” case. Done. Included (Fifth line in Introduction section). Moreover, I suggest adding a sentence regarding the role of animal reservoirs and their distribution that contributes (as the presence of vectors) in the maintenance of the transmission cycle. Done, now included (Final sentence in Introduction section). In the Methods:You could clarify which and/or how many countries have been used as keyword for the search pipeline Now included (In Methods section after keyword specification). Figure 6: In the caption: the citation trends is from Web of Science (as reported in the results), not Scopus. Corrected." } ] } ]	1	https://f1000research.com/articles/7-1987
https://f1000research.com/articles/8-280/v1	12 Mar 19	{ "type": "Research Note", "title": "Causes of reporting bias: a theoretical framework", "authors": [ "Jenny T van der Steen", "Gerben ter Riet", "Cornelis A van den Bogert", "Lex M Bouter", "Gerben ter Riet", "Cornelis A van den Bogert", "Lex M Bouter" ], "abstract": "Reporting of research findings is often selective. This threatens the validity of the published body of knowledge if the decision to report depends on the nature of the results. Studies on causes and mechanisms underlying selective reporting may help to avoid or reduce reporting bias. Such research should be guided by a theoretical framework of possible causal pathways that lead to reporting bias. We build upon a classification of determinants of selective reporting that we recently developed in a systematic review of the topic. The resulting theoretical framework features four clusters of causes. There are two clusters of necessary causes: (A) motivations (e.g. a preference for positive findings) and (B) means (e.g. a flexible study design). These two combined represent a sufficient cause for reporting bias to occur. The framework also features two clusters of component causes: (C) conflicts and balancing of interests referring to the individual or the team, and (D) pressures from science and society. The component causes may modify the effect of the necessary causes or may lead to reporting bias mediated through the necessary causes. Our theoretical framework is meant to inspire further research and to create awareness among researchers and end-users of research about reporting bias and its causes.", "keywords": [ "Causality", "publication bias", "questionable research practice", "reporting bias", "research design", "selective reporting" ], "content": "Background\n\nSelective reporting of research findings presents a large-scale problem in science, substantially affecting the validity of the published body of knowledge (Bouter et al., 2016; Dwan et al., 2014; van den Bogert et al., 2017). Reporting bias (publication bias or outcome reporting bias) occurs when the decision to report depends on the direction or magnitude of the findings. In clinical research, registration of trials prior to data collection is used to prevent selective reporting with some success (Chan et al., 2017; Gopal et al., 2018). However, it is insufficiently effective because despite registration or publication of the study protocol, trial results often remain partially or completely unpublished (Jones et al., 2013) and selective reporting of “positive findings” also occurs among trials registered at, for example, clinicaltrials.gov (Dechartres et al., 2016).\n\nAlthough many epidemiological studies have described the occurrence of selective reporting, very few studies have targeted its causes. In particular there is little high-quality evidence on effective interventions. To develop effective interventions against reporting bias, we need a good understanding of possible contributions of actors involved (such as academic environment, editors, researchers) and of possible mechanisms. We also need clear hypotheses of how causes may be interrelated and how actors are involved.\n\nWe recently developed a taxonomy of putative determinants of selective reporting abstracted from the literature. We used qualitative content analyses of empirical and non-empirical studies until we reached saturation, which indicates that the categories likely cover all important putative determinants of selective reporting. This resulted in 12 categories (Table 1).\n\n(Modified from the taxonomy of determinants presented in Table 3 in: Determinants of selective reporting: A taxonomy based on content analysis of a random selection of the literature. van der Steen JT et al. PLoS One. 2018 Feb 5;13(2):e0188247. doi: 10.1371/journal.pone.0188247.)\n\n* With study design, we mean broader design issues than just type of research design, including also definitions, outcomes, analytic plans etc.\n\nIn the literature review we also found some instances of hypothesized effect modification of the determinants of selective reporting, so that the effects of determinants are assumed not to be simply additive. For example, “Outcomes could be deemed post hoc to have little clinical relevance if they fail to show significant findings and may thus be omitted when accommodating space limitations” (Chan & Altman, 2005). In this case, a preference, namely statistically significant findings, combined with editorial practices lead to reporting bias. Similarly, Ioannidis (2005) hypothesized that a focus on preferred, positive findings could result in reporting of non-reproducible findings (only) if there is also an opportunity to do so through flexibility in study designs and freedom in reporting on it. That is, he concludes that “The greater the flexibility in designs, definitions, outcomes, and analytical modes in a scientific field, the less likely the research findings are to be true” because “Flexibility increases the potential for transforming what would be ‘negative’ results into ‘positive’ results.“\n\n\nA framework of possible causal pathways to reporting bias\n\nAlong these lines, we hypothesize that the combination of two of the most common categories in our review (van der Steen et al., 2018) –– i.e., focusing on preferred findings and employing a poor or flexible study design, suffices to cause bias through selective reporting. Inspired by Rothman’s (1976) framework of necessary, sufficient and component causes, through multiple discussions, we inductively derived Figure 1 which shows clusters covering these and the ten other categories of determinants and their possible interrelationships. The two categories are part of clusters A (motivations) and B (means). We view both clusters A and B as necessary causes, that is, they are both part of any sufficient cause of reporting bias. There is also effect modification between A and B because reporting bias is not possible with A or B alone. Note that the preference need not be authors’ preference; it may also be that of a reviewer or editor. In addition to clusters A and B, we propose clusters C and D containing categories of component causes which are discussed in the next section.\n\nBullet points indicate the 12 categories of determinants of selective reporting subsumed under four higher-level clusters A, B, C, and D. Note that the figure implies effect modification between A and B (necessary causes) because there will be no reporting bias with A or B alone. Effect modification (“X”) may also occur by C or D and thus make the joint effect of A and B stronger. Mediation (“M”) may occur if the necessary causes (A and B) mediate the effect of D. Mediation may also occur if C mediates the effects of D on A and B, which in its turn leads to reporting bias.\n\nPoor or flexible study design may offer the means for selective reporting in addition to limitations in reporting and editorial practices (cluster B in Figure 1). In parallel, we placed “prejudice” in cluster A together with “preference for particular findings” because both may, whether consciously or not, represent a motivation for behaviour that leads to reporting bias. The possible motivations, wishes and beliefs in cluster A are different concepts that may result in “wishful thinking” (Bastardi et al., 2011) and in motivated reasoning around the interpretation of scientific findings (e.g. to serve political interests; Colombo et al., 2016; Kraft et al., 2015). Persons may or may not be fully aware of their motivations and the resulting behaviour may or may not be intentional (Greenland, 2009). At the root of reporting bias may thus lay a basic human attitude, the very natural tendency to make public our successes (Dickersin & Min, 1993).\n\nThe pertinence of the second necessary cause (cluster B)––multiple opportunities to select what to analyse or report––is illustrated by the many degrees of freedom that researchers have but should not be tempted to use (in performing psychological research: Wicherts et al., 2016). The necessary causes thus represent having a motive (preference or prejudice; cluster A) and the means (opportunities in study design or reporting; cluster B). Together they form a sufficient cause for reporting bias.\n\nObviously, researchers and editors are key stakeholders because commonly they decide what is actually being reported and what is not. It can be argued that researchers are the most important because a single editor’s decision is not decisive for non-publication or selective publication. Researchers are actors in three of the four categories in clusters A and B that represent the necessary causes, while editors are key players in only one category (in cluster B; Figure 1). Note that we assume actors in the field are capable.\n\nAfter a series of rejections researchers may doubt whether reporting is worth the effort under the pressure of lack of resources such as time. Balancing effort and output is placed in cluster C (component cause conflicts and balancing of interests; Figure 1). Cluster C also includes relationship and collaboration issues and dependence upon sponsors. Cluster C thus represents conflicts of interests, individuals and teams juggling with harmony in relationships and time investments.\n\nOther component causes represent pressures from science and society (cluster D), so from the wider environment. The individual researcher has less control over type C, and in particular type D causes, than over motivations (A) and means (B). C and D cannot fully control or explain individuals’ decisions, but they may shape motivations (A) and means (B). When this is the case the effect on reporting bias of the categories in cluster C or D is mediated through the categories contained in cluster A or B. For example, important news is selectively reported but what is deemed important news is shaped by the development within a scientific domain (cluster C; Preston et al., 2004). Also, researchers’ collaborations or relations with sponsors may nudge them to selectively report the preferences of others. A final example is academic publication system hurdles (cluster D) and dependence upon sponsors (cluster C) leading to reporting bias through their impact on the combination of a preference for positive findings and the opportunities that flexible designs offer.\n\n\nDiscussion\n\nWe propose a theoretical framework of reporting bias by relating and ordering 12 determinant categories that we derived from the literature (van der Steen et al., 2018). We further combined these categories in four clusters (A-D).\n\nThe model has more layers and is more refined than we anticipated when we wrote a protocol to develop a taxonomy of determinants of selective reporting and their interrelationships. We then expected a central role for preferences for particular “positive” findings only (van der Steen et al., 2018 Supplement 1, Figure 1). However, having the means is necessary as well. Although the determinants in our model are mostly based on research in the biomedical area, the model fits well with the “Desire-Belief-Opportunity” (DBO) model that analytical sociologists use to explain various phenomena (Hedström, 2005) and which we came across after developing our theoretical framework. Desire and Belief concur with the two motivations in cluster A, while opportunities (alternative actions available to the actor) represent the means in cluster B.\n\nTheory may guide the development of interventions as research often does not systematically consider contextual and individual factors that influence delivery of an intervention. Thus, theory may help avoid an ad hoc or data-driven approach to attempts to reduce reporting bias. Although one might assume that interventions addressing reporting bias effectively will be complex, the removal of a single necessary cause is obviously effective. For example, a potentially very effective measure that funders and (medical) ethics committees could adopt is systematic monitoring of all written research outputs and comparing the outcomes reported therein to the corresponding research protocols and statistical analysis plans and eventual amendments. This would require that these organizations make submission to them or to a publicly available repositories mandatory. The approach would become practically feasible if software comparing protocols to publications becomes available (ter Riet & Bouter, 2016). In the jargon of this paper, this approach would eliminate the necessary cause ‘means.’ Given suitable negative reinforcements (punishments) following incomplete reporting, such measures may also reduce motivation to report selectively. Similarly, elements from the component causes contained in cluster C and D that are highly prevalent and strongly modify the combined effect of cluster A and B may be prioritized targets. Mediators can also be good candidates for intervention. For example, component causes contained in cluster C may mediate the impact of elements of D on elements of clusters A or B.\n\nIn addition to informing the development of interventions that are subsequently evaluated, our framework may also help to identify high risk scientific fields. For example, areas where designs offer considerable flexibility or where the researchers’ degrees of freedom are combined with strong beliefs or a mission to disseminate particular outcomes. Of course, based on research, our theoretical framework may need to be adapted. Motive and means may be stable clusters but the C and D type causes may change as science changes. Future work may also help to refine the framework’s relevance for specific disciplinary fields (e.g., non-clinical biomedical research). Nevertheless, because the causal pathways seem plausible, were derived from the literature on selective reporting and is congruent with theory developed in the social sciences (Hedström, 2005), we feel that the current work can already help to design further research on the effectiveness of interventions.\n\n\nData availability\n\nPLOS ONE Supplement 2 to article van der Steen et al., 2018. Determinants of selective reporting abstracted from the selected literature. “S2 File. Dataset with determinants.” In Excel available from: https://doi.org/10.1371/journal.pone.0188247.s003 (van der Steen et al., 2018)\n\nPLOS ONE Supplement 3 to article van der Steen et al., 2018. Categories of determinants of selective reporting with literature references. “S3 file. References to the 64 articles included in the determinant analysis, per category.” In Word available from: https://doi.org/10.1371/journal.pone.0188247.s004 (van der Steen et al., 2018)\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).", "appendix": "Grant information\n\nThe writing of the article was supported by personal grants to JTvdS from the Netherlands Organisation for Scientific Research (NWO; Innovational Research Incentives Scheme: Vidi grant number 917.11.339), and from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (Consolidator grant agreement No 771483), and by Leiden University Medical Center, Leiden, The Netherlands.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nBastardi A, Uhlmann EL, Ross L: Wishful thinking: belief, desire, and the motivated evaluation of scientific evidence. Psychol Sci. 2011; 22(6): 731–2. PubMed Abstract \| Publisher Full Text\n\nBouter LM, Tijdink J, Axelsen N, et al.: Ranking major and minor research misbehaviors: Results from a survey among participants of four World Conferences on Research Integrity. Res Integr Peer Rev. 2016; 1: 17. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChan AW, Altman DG: Identifying outcome reporting bias in randomised trials on PubMed: review of publications and survey of authors. BMJ. 2005; 330(7494): 753. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChan AW, Pello A, Kitchen J, et al.: Association of trial registration with reporting of primary outcomes in protocols and publications. JAMA. 2017; 318(17): 1709–1711. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nColombo M, Bucher L, Inbar Y: Explanatory Judgment, Moral Offense and Value-Free Science. Rev Philos Psychol. 2016; 7(4): 743–63. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDechartres A, Bond EG, Scheer J, et al.: Reporting of statistically significant results at ClinicalTrials.gov for completed superiority randomized controlled trials. BMC Med. 2016; 14(1): 192. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDickersin K, Min YI: NIH clinical trials and publication bias. Online J Curr Clin Trials. 1993; Doc No 50. PubMed Abstract\n\nDwan K, Altman DG, Clarke M, et al.: Evidence for the selective reporting of analyses and discrepancies in clinical trials: A systematic review of cohort studies of clinical trials. PLoS Med. 2014; 11(6): e1001666. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGopal AD, Wallach JD, Aminawung JA, et al.: Adherence to the International Committee of Medical Journal Editors' (ICMJE) prospective registration policy and implications for outcome integrity: a cross-sectional analysis of trials published in high-impact specialty society journals. Trials. 2018; 19(1): 448. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGreenland S: Accounting for uncertainty about investigator bias: disclosure is informative. J Epidemiol Community Health. 2009; 63(8): 593–8. PubMed Abstract \| Publisher Full Text\n\nHedström P: Dissecting the Social - On the Principles of Analytical Sociology. Cambridge Cambridge University Press. 2005. Reference Source\n\nIoannidis JP: Why most published research findings are false. PLoS Med. 2005; 2(8): e124. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJones CW, Handler L, Crowell KE, et al.: Non-publication of large randomized clinical trials: Cross sectional analysis. BMJ. 2013; 347: f6104. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKraft PW, Lodge M, Taber CS: Why People \"Don't Trust the Evidence\": Motivated Reasoning and Scientific Beliefs. Ann Am Acad Polit Soc Sci. 2015; 658(1): 121–33. Publisher Full Text\n\nPreston C, Ashby D, Smyth R: Adjusting for publication bias: modelling the selection process. J Eval Clin Pract. 2004; 10(2): 313–22. PubMed Abstract \| Publisher Full Text\n\nRothman KJ: Causes. Am J Epidemiol. 1976; 104(6): 587–92. PubMed Abstract \| Publisher Full Text\n\nter Riet G, Bouter LM: How to end selective reporting in animal research. In Animal models for human cancer: discovery and development of novel therapeutics. Wiley: Weinheim. ISBN 9783527339976. 2016. Publisher Full Text\n\nvan den Bogert CA, Souverein PC, Brekelmans CT, et al.: Primary endpoint discrepancies were found in one in ten clinical drug trials. Results of an inception cohort study. J Clin Epidemiol. 2017; 89: 199–208. PubMed Abstract \| Publisher Full Text\n\nvan der Steen JT, van den Bogert CA, van Soest-Poortvliet MC, et al.: Determinants of selective reporting: A taxonomy based on content analysis of a random selection of the literature. PLoS One. 2018; 13(2): e0188247. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWicherts JM, Veldkamp CL, Augusteijn HE, et al.: Degrees of freedom in planning, running, analyzing, and reporting psychological studies: A checklist to avoid p-hacking. Front Psychol. 2016; 7: 1832. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "46634", "date": "11 Apr 2019", "name": "Ksenija Bazdaric", "expertise": [ "Reviewer Expertise research ethics", "psychology", "medical informatics" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI was happy to review a manuscript about a theoretical framework in the field of reporting bias. I think the authors have proposed an interesting perspective but my major remark is that they try to explain human behaviour with an epidemiological model for which I don’t find a body of evidence in the literature that could convince me.\n\nComments:\n\nBackground: In clinical research, registration of trials prior to data collection is used to prevent selective reporting with some success – please delete “some success” because it is further explained. \"A framework of possible causal pathways to reporting bias - Motivations and means. Along these lines, we hypothesize that the combination of two of the most common categories in our review (van der Steen et al., 2018) –– i.e., focusing on preferred findings and employing a poor or flexible study design, suffices to cause bias through selective reporting...\" – how do you then comment on the replication crisis in psychology and the experiments that were replicated in the same laboratories? They were motivated to replicate and for sure were not sloppy. Do we have a poor designed field here or are there other factors? I would like a more detailed explanation1,2,3. A theoretical framework for reporting bias. Rothman’s theoretical model – is there any evidence in practice for this model in relation to human behaviour? Figure 1 and the model: It is an interesting figure, but the same could be explained by some other general theories, for example 'Theory of planned behaviour' (of course evidence cannot be confirmed as we have a replication crisis in psychology). I don’t believe human behaviour can be explained with an epidemiological model although it is very nice. Also, the model itself does not have a word about ethical climate and other possible external factors. Why did you exclude them? Do you consider them stable in all environments? There is some evidence that peers can predict replication2? How do you comment? Could you include some external factors in your model? Like environment, ethical climate, etc… The sentence: \"At the root of reporting bias may thus lay a basic human attitude, the very natural tendency to make public our successes\" – this is not clear at all. At a root of everything probably lies personality and attitude, but I don’t understand the meaning of the sentence here. Obviously, researchers and editors are key stakeholders because commonly they decide what is actually being reported and what is not. – I would say that sometimes we cannot report everything because we have only 3000-4000 words (here is the role of editors). After a series of rejections researchers may doubt whether reporting is worth the effort under the pressure of lack of resources such as time. I advise you to read a case study from Helene Speyer - The value of a “failed” trial4.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Not applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? No source data required\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4708", "date": "17 Jul 2019", "name": "Jenny van der Steen", "role": "Author Response", "response": "I was happy to review a manuscript about a theoretical framework in the field of reporting bias. I think the authors have proposed an interesting perspective but my major remark is that they try to explain human behaviour with an epidemiological model for which I don’t find a body of evidence in the literature that could convince me. Reponse: Thank you so much for reviewing our manuscript and considering it from the perspective of research in humanities.Thank you also for your interesting viewpoint regarding types of models, which is helpful to clarify the background of our model.We respectfully disagree that qualifications such as epidemiological model, or behavioural model apply well to our model. The model has been developed from determinants of selective reporting mostly in the medical and biomedical literature, but we also retrieved determinants from, e.g., the humanities (11% of articles, van der Steen et al., PLOS One 2018). We analyzed the determinants inductively using qualitative content analyses. This means we did not impose any previous model when we identified and categorized the determinants. Texts from the original articles inspired us to consider how determinants might work together, also distinguishing between clusters of related determinants. In a next step, we indeed recognized and used epidemiological terms to label relationships between categories of determinants in terms of effect modification and interaction. Text in the original articles also inspired us to consider whether some categories of determinants could be necessary causes, which is also terminology used in epidemiology.However, we could just not use the terms and arrive at the same model inspired by the literature about how categories of determinants could work together. We feel that using certain terminology does not suffice to qualify the model in such a way that it limits application to fields that do not use that terminology. The essence is our inductive way of forming categories and in part, also the thinking of relationships has been taken from the literature using an open approach about how categories may work together.Moreover, the outcome is reporting bias in the literature which is very broad. It cannot be brought about by the behavior of a single individual. Therefore, we do not claim we can predict individual behavior. In contrast, the core of our model ¾ the necessary causes that indeed refer the most to behavior of individuals ¾ probably resembles the most a broad model from sociology (“Desire-Belief-Opportunity,” DBO), which we found by coincidence after having conceived our model. Finally, the cluster of Pressures from science and society comprises four societal determinants of reporting bias and we visualized it can impact behavior of individuals directly but also indirectly.In all, we believe there is little reason to qualify the model as belonging to a single particular discipline imposing disciplinary boundaries around its application. We rather hope that our model will cross such boundaries and help researchers from different disciplines and perhaps also policy makers to work together to address reporting bias. In the Background paragraph of “Basis for a theoretical causal framework: hypothesized determinants of selective reporting and their interrelationships” we added that we had used an inductive approach to develop the categories. Further, we connected the paragraph with the next paragraph (“A framework of possible causal pathways to reporting bias”) more clearly by starting it with “Based on what we found in the literature.” Next, we reversed the order in the sentence about how we applied Rothman’s classification of causes, because indeed we first agreed on the relationships and we used Rothman’s labels to express the relationships in terms of his nomenclature. The sentence that started with “Inspired by Rothman’s (1976) framework of necessary, sufficient and component causes,” now starts with “Through multiple discussions, we inductively derived Figure 1.” In the following sentence, we also clarified that we applied terms from epidemiology to what we found in an inductive way, by starting it with: “Applying more epidemiological terms to the generic model we developed,” We reflect on the approach and background of the authors by adding to the same paragraph that the discussions were “among the team with experience in both qualitative and quantitative research.” Finally, to the first sentence of the Discussion, we added that our theory is “broad” and that we combined categories into clusters “inductively”, “using existing epidemiologic terminology to label relationships.”Comments: 1. Background: In clinical research, registration of trials prior to data collection is used to prevent selective reporting with some success – please delete “some success” because it is further explained.Response: We deleted as suggested “some success.”2. \"A framework of possible causal pathways to reporting bias - Motivations and means.Along these lines, we hypothesize that the combination of two of the most common categories in our review (van der Steen et al., 2018) –– i.e., focusing on preferred findings and employing a poor or flexible study design, suffices to cause bias through selective reporting...\" – how do you then comment on the replication crisis in psychology and the experiments that were replicated in the same laboratories? They were motivated to replicate and for sure were not sloppy. Do we have a poor designed field here or are there other factors? I would like a more detailed explanation1,2,3.Response: Thank you for your comment. The interesting Nature survey on non-reproducibility that you cited (Baker, 2016)1 indicates that researchers believe selective reporting is a major contributor to research often not being replicable, and also a number of factors which in our model are determinants of reporting bias (e.g., pressure to publish and poor experimental design). Of course the Nature survey did not aim at offering a model or mechanisms and therefore outcomes and determinants are not being distinguished. Despite probably sharing determinants, the survey also includes “methods, code unavailable” as a reason for non-reproducibility which underlines that reporting bias is not the only reason that research is not reproducible.Further, perhaps it was not clear enough that the model does not assert that if means and motivations are there, these are sufficient causes but this will not necessarily result in reporting bias. The model includes Conflicts and balancing of interest and Pressures from science and society which can modify effects on reporting bias. Further, not included in our model are interventions. Interventions may alter motivations or change the flexibility of designs and therefore diminish effects on reporting bias. We hope the model will inspire development of targeted interventions.After “We view both clusters A and B as necessary causes, that is, they are both part of any sufficient cause of reporting bias” we added: “This does not mean that reporting bias will always be the result of presence of A and B because effects can be mitigated by interventions and modified by component causes.” Further, we added in this paragraph the term “may” in: “Together they may form a sufficient cause for reporting bias.”The model does not specify what kind of motivations researchers may have, what would be the most interesting outcome in a particular case (e.g., in our review we found motivations to try to find the same as other research early in development of a field, but motivations later on changed to find deviations). Also we do not provide criteria for how much or how little flexibility in design would be optimal or qualify particular research as sloppy.We further extended the paragraph with a citation from our review about motivation to report news and how the contents of the news may vary over time, in addition to citing it later on referring to development of a field: “Success can be defined in different, or even opposite ways as suggested by Rosenthal and Rubin cited by Preston et al. (2004) which was part of our review: “early in the history of a research domain results in either direction are important news but that later, when the preponderance of evidence has supported one direction, significant reversals are often more important news than further replications.” This example also illustrates the intertwining of individual motivations and societal factors.Flexibility of design may actually be an advantage in some cases as long as the reporting about the methods is transparent, and some designs are flexible by nature. Such specifications would require empirical research.To the last paragraph of the Discussion, we added: “Further empirical research is needed to specify, for example, what the optimal level of flexibility for a particular field and study design would be.”3. A theoretical framework for reporting bias. Rothman’s theoretical model – is there any evidence in practice for this model in relation to human behaviour? Figure 1 and the model: It is an interesting figure, but the same could be explained by some other general theories, for example 'Theory of planned behaviour' (of course evidence cannot be confirmed as we have a replication crisis in psychology). I don’t believe human behaviour can be explained with an epidemiological model although it is very nice. Also, the model itself does not have a word about ethical climate and other possible external factors. Why did you exclude them? Do you consider them stable in all environments?There is some evidence that peers can predict replication2? How do you comment? Could you include some external factors in your model? Like environment, ethical climate, etc…Response: As explained above in our response to the general comment, Rothman’s terms were applied later on, when the categories of determinants were already retrieved in an inductive manner from literature about determinants of selective reporting. We believe that the theory of planned behavior is not very suitable to explain reporting bias as a phenomenon based on behavior of multiple individuals, teams, institutes, peer groups, or society. How would it incorporate the categories we found in the literature, such as “Unfavourable geographical or regulatory environment” (Table 1)? Further, various other theories may apply to parts of our model, for example, from criminology, theories about strain and goal displacement.Reviewing our dataset with 497 determinants, we find, for example, one of the determinants from the literature classified under this heading was “Increased accountability from medical journals regarding ethics and conflicts of interest” (Dieterich et al., 2013; Supplemental file 2 to the review in PLOS One). It relates to the ethical climate in the publishing enterprise. Of note, at this point, we would not revise categories that are not being supported by the content analysis of the literature, but we can clarify the existing categories by providing more examples.To further clarify the category of Unfavourable geographical or regulatory environment” in Table 1, we added the example of “ethics in publishing enterprise”Replication studies can help define the scope of problems with replicability which may be due to reporting bias only in part. Replication studies can also affect the four clusters A-D. It can impact Motivations (e.g., researchers becoming more interested in study results that are likely replicated or researchers conducting replication studies being more open to, or working towards in null results), Means (e.g. rise of specific journals that support publishing replicated studies), Conflicts and balancing of interests (e.g. earmarked resources for replication studies becoming available), and Pressures from science and society (e.g. less creative and innovative but rigorous research becoming more salonfähig). We added this example of replication studies impacting clusters A-D to the Discussion.4. The sentence: \"At the root of reporting bias may thus lay a basic human attitude, the very natural tendency to make public our successes\" – this is not clear at all. At a root of everything probably lies personality and attitude, but I don’t understand the meaning of the sentence here.Response: we cited from an article included in the review in PLOS One because those articles provided the data we analysed to develop the determinant categories. Perhaps the reasoning was not clear enough.Although we would not repeat the methods of the review in full, we added explicit reference to the review earlier in the text, inserting it in the second paragraph of the Background, after ”We recently developed a taxonomy of putative determinants of selective reporting abstracted from the literature” (van der Steen et al., 2018).Further, the authors of the statement find a general attitude to show success relevant for reporting bias. They do not claim to explain all human behavior. We feel the emphasis about showing success is a key issue. We deleted “a basic human attitude” to avoid confusion about the key point.5. Obviously, researchers and editors are key stakeholders because commonly they decide what is actually being reported and what is not. – I would say that sometimes we cannot report everything because we have only 3000-4000 words (here is the role of editors).Response: We agree. We found this in several articles. Table 1 therefore shows that one of twelve categories of determinants was “Limitations in reporting and editorial practices” that was explained as “Constraints and barriers to the practice of reporting relevant detail” with included among the examples “Journal space restrictions” Further, academic publication system hurdles is a separate category at the level of cluster D, science and society and it also includes editors as main actors.6. After a series of rejections researchers may doubt whether reporting is worth the effort under the pressure of lack of resources such as time. I advise you to read a case study from Helene Speyer - The value of a “failed” trial4.Response: There is a point where efforts to try to publish work are not in balance with the perceived or witnessed chances of it actually being published. This was what we found in the literature as one of the determinants of selective reporting.To clarify the basis of the statement, We started the second section of “A framework of possible causal pathways to reporting bias,” with: “In the review, we found that..”Obviously, trials that fail because they do not confirm the hypothesis are as valuable as trials with expected results or results hoped for. We appreciate the article of Speyer, her personal account of initial doubts and efforts to publish the results of a strong trial that in fact delivered strong evidence that the intervention was ineffective. It calls for editors and reviewers to reflect on their own possible biases and focus on the quality of the work rather than judging the quality of the work in light of their perspectives on the results and how the results may or may not differ from previous work.We added to the Discussion: The model also shows that for example editors may influence the outcome in multiple ways; first, directly via Means (the category Limitations in reporting and editorial practices). Second, editors as a collective affect the rejection rates through Academic publication system hurdles (the category of Pressures from science and society), but also the extent to which authors find efforts to publish worthwhile (category of cluster Conflicts and balancing of interests).The latter is illustrated by the personal account of Speyer (2018). Other changes.We edited the text at several places, shortening it where possible.In the Discussion, we shortened the example of an intervention with regard to the means for comparing protocols and publication. Software to compare protocol and publication was not necessary in the NIHR Journal Library system and we added reference to Wright et al. (2018) for this." } ] }, { "id": "47229", "date": "20 May 2019", "name": "Arnaud Vaganay", "expertise": [ "Reviewer Expertise Meta-research", "research integrity", "systematic reviews", "social sciences", "science policy", "education", "experimental and quasi-experimental methods." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI was pleased to review this manuscript. I assessed the proposed theoretical framework based on the following criteria: (1) utility, (2) comprehensiveness, (3) parsimony, (4) testability, (5) heurism, and (6) scope. Ironically, these criteria are not part of an established theoretical framework; they only reflect a brief review of the literature on the subject.\nI would rate the utility of the proposed framework as high. As far as I know, this is one of the first attempts to synthesize the literature on the factors driving reporting bias. Additionally, and as mentioned by the authors, understanding these factors is essential to the development of mitigating measures.\n\nI would rate the heurism (i.e. evidence base) of the framework as low. Granted, there is now a large body of literature on the prevalence of reporting bias and on the possible factors driving it. However, virtually none of these studies are experimental, so all the relationships found between the occurrence of reporting bias and the 12 categories are at best correlational – not causal. This is, in my view, the most fundamental flaw of the framework. The good news is, it could be easily addressed by changing the title of the manuscript from “Causes of reporting bias” to “Drivers of reporting bias”.\n\nRelated to the above-mentioned point, I would rate the testability of the framework as low. The only testable driver of reporting bias is “financial conflict of interests” (FCOI). Most other drivers (prejudice, relationship and collaboration issues, doubts about reporting being worth the effort, etc.) would be hard to test empirically. Virtually none of these drivers can be tested in experimental conditions.\n\nI was unable to rate the comprehensiveness of the framework for three reasons. Firstly, and as already mentioned, I do not think that the proposed framework can realistically identify the possible “causes” of reporting bias – let alone all the possible causes. Secondly, the authors constructed the typology based on a “systematic review”, which I haven’t seen. Systematic reviews of narrowly defined questions are notoriously hard to conduct; I don’t think that a review of such a broad question can be truly systematic. Thirdly, I don’t have enough knowledge of the topic to suggest missing/alternative categories.\n\nI would rate the parsimony of the framework as medium. On the one hand, I do not think it is possible to propose a parsimonious framework when the scope of the theory is so broad. On the other hand, Table 1 is much more parsimonious than the vast literature it draws on.\n\nI would rate the scope of the framework as fairly high. The authors acknowledge that the determinants of the model “are mostly based on research in the biomedical area”. However, I would argue that the 12 categories listed in the framework are also relevant to social research, which is the type of research that I do myself.\n\nIn conclusion, I am grateful for the authors’ contribution to the literature and do believe that the proposed framework could help the research community address the problem of reporting bias. However, I have some concerns regarding (i) the strength of the evidence used by the authors to make causal claims; and (ii) the testability of the framework.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Not applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? No source data required\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4709", "date": "17 Jul 2019", "name": "Jenny van der Steen", "role": "Author Response", "response": "1. I was pleased to review this manuscript. I assessed the proposed theoretical framework based on the following criteria: (1) utility, (2) comprehensiveness, (3) parsimony, (4) testability, (5) heurism, and (6) scope. Ironically, these criteria are not part of an established theoretical framework; they only reflect a brief review of the literature on the subject.Response: Thank you. We appreciate the reviewing of our work against a series of relevant criteria.2. I would rate the utility of the proposed framework as high. As far as I know, this is one of the first attempts to synthesize the literature on the factors driving reporting bias. Additionally, and as mentioned by the authors, understanding these factors is essential to the development of mitigating measures.Response: Thank you, indeed we aimed at identifying candidate determinants of reporting bias and interrelationships with a view to develop interventions in a more systematic way, and to allow for testing to improve the theoretical framework. 3. I would rate the heurism (i.e. evidence base) of the framework as low. Granted, there is now a large body of literature on the prevalence of reporting bias and on the possible factors driving it. However, virtually none of these studies are experimental, so all the relationships found between the occurrence of reporting bias and the 12 categories are at best correlational – not causal. This is, in my view, the most fundamental flaw of the framework. The good news is, it could be easily addressed by changing the title of the manuscript from “Causes of reporting bias” to “Drivers of reporting bias”. Response: We fully agree that the evidence base for the theoretical framework is still low. We found only 1 randomized trial in the 64 articles we analysed in the review. Numerous studies have been performed, and also interventions such as training and pre-registration have been implemented but very few interventions have been evaluated thoroughly. This is why we speak about putative determinants and plausible mechanisms. We are not sure whether “drivers” is essentially different from “causes.” Heurism is probably in how we could learn about causes. To this end, it is important to consider the subtitle which is “a theoretical framework.” This can be developed also on the basis of a modest evidence base, because the aim is to test the model’s assumptions and adapt the model if necessary. This might progress the field at a faster pace than continuing without any or only an implicit theoretical framework. At least, we believe that considering the current evidence base, developing a theoretical framework is timely. We do recognize the small evidence base of the framework. To the last paragraph of the Discussion, we added a sentence, to start the paragraph with explicitly recognizing the small evidence base: “Currently, the evidence for the theoretical framework is limited.” 4. Related to the above-mentioned point, I would rate the testability of the framework as low. The only testable driver of reporting bias is “financial conflict of interests” (FCOI). Most other drivers (prejudice, relationship and collaboration issues, doubts about reporting being worth the effort, etc.) would be hard to test empirically. Virtually none of these drivers can be tested in experimental conditions. Response: We agree that testing the model will require thoughtful consideration of putative determinants and how these could be manipulated. Models can be helpful even if not testable in a traditional way and many theories have been developed and maintained this way. Further, we need to consider how other than experimental designs can facilitate a better understanding of causes (importantly, more in-depth and thorough qualitative research which has been largely neglected in this area). The model itself shows that the issue is multifactorial which means often multicomponent interventions are required which complicates identifying contributions of single components. Creative experimental designs such as manipulating realistic scenarios withholding or adding information in different conditions and choice experiments may offer alternative ways of researching researchers in situations that are otherwise standardized. We agree that testing the model will require huge efforts, but we do not believe it is impossible.Perhaps the focus of the Discussion was too much on experimental work. The model may also improve the quality of design of observational studies. For example, when, to assess associations with a specific determinants, the model inspires to measure important possible confounding factors. We added to the Discussion: “The model may also assist in assessing potential confounding factors in observational work in which associations between a specific determinant and reporting bias is assessed.” 5. I was unable to rate the comprehensiveness of the framework for three reasons. Firstly, and as already mentioned, I do not think that the proposed framework can realistically identify the possible “causes” of reporting bias – let alone all the possible causes. Secondly, the authors constructed the typology based on a “systematic review”, which I haven’t seen. Systematic reviews of narrowly defined questions are notoriously hard to conduct; I don’t think that a review of such a broad question can be truly systematic. Thirdly, I don’t have enough knowledge of the topic to suggest missing/alternative categories.Response: Our systematic review was published in PLOS One (2018). We apologize that we had referred to it only in the section explaining the model and not yet in the Background section. The search was systematic, but we analysed only a random sample of the literature because indeed the broad question did not allow to analyze all relevant literature. However, we assessed saturation of the qualitative content analyses of determinants abstracted from the literature in two ways, prospectively and retrospectively, both of which indicated that finding additional categories of determinants through analyzing more articles was unlikely. We were interested in the main categories which cover a number of specific determinants which may also differ somewhat for different disciplines. We commented on possible refinement needed for specific disciplinary fields in the Discussion section.We added explicit reference to the review, inserting it after ”We recently developed a taxonomy of putative determinants of selective reporting abstracted from the literature” (van der Steen et al., 2018). Next, we clarified the methods by several additions to the paragraph of “A framework of possible causal pathways to reporting bias.” 6. I would rate the parsimony of the framework as medium. On the one hand, I do not think it is possible to propose a parsimonious framework when the scope of the theory is so broad. On the other hand, Table 1 is much more parsimonious than the vast literature it draws on.Response: Thank you. In the qualitative content analyses of the literature we aimed at grouping together related determinants. 7. I would rate the scope of the framework as fairly high. The authors acknowledge that the determinants of the model “are mostly based on research in the biomedical area”. However, I would argue that the 12 categories listed in the framework are also relevant to social research, which is the type of research that I do myself. Response: Thank you. The literature we analysed predominantly concerned clinical medicine and biomedicine, but the humanities represented 11% of articles. The model also resembles the Desire-Belief-Opportunity” (DBO) model from sociology. 8. In conclusion, I am grateful for the authors’ contribution to the literature and do believe that the proposed framework could help the research community address the problem of reporting bias. However, I have some concerns regarding (i) the strength of the evidence used by the authors to make causal claims; and (ii) the testability of the framework.Response: We hope that our theory about causality which clearly cannot reach farther than the evidence base and literature so far, inspires researchers to develop smart designs to discover causal mechanisms and to propose adaptation of the theory if needed. Thank you for this interesting and thoughtful input. Other changes.We edited the text at several places, shortening it where possible.In the Discussion, we shortened the example of an intervention with regard to the means for comparing protocols and publication. Software to compare protocol and publication was not necessary in the NIHR Journal Library system and we added reference to Wright et al. (2018) for this." } ] } ]	1	https://f1000research.com/articles/8-280
https://f1000research.com/articles/8-2/v1	02 Jan 19	{ "type": "Case Report", "title": "Case Report: Ocular toxoplasmosis in a WHIM syndrome immunodeficiency patient", "authors": [ "David H. McDermott", "Lauren E. Heusinkveld", "Wadih M. Zein", "H. Nida Sen", "Martha M. Marquesen", "Mark Parta", "Sergio D. Rosenzweig", "Gary A. Fahle", "Michael D. Keller", "Henry E. Wiley", "Philip M. Murphy", "David H. McDermott", "Lauren E. Heusinkveld", "Wadih M. Zein", "H. Nida Sen", "Martha M. Marquesen", "Mark Parta", "Sergio D. Rosenzweig", "Gary A. Fahle", "Michael D. Keller", "Henry E. Wiley" ], "abstract": "A patient with WHIM syndrome immunodeficiency presented with sudden painless right eye blindness associated with advanced retinal and optic nerve damage. Toxoplasma gondii was detected by PCR in vitreous fluid but not serum. The patient was treated with trimethoprim-sulfamethoxazole. Vision did not return; however, the infection did not spread to involve other sites. Toxoplasmosis is rare in primary immunodeficiency disorders and is the first protozoan infection reported in WHIM syndrome.", "keywords": [ "Toxoplasma gondii", "Treatment", "Retinitis", "Optic neuritis", "Genetic Immunodeficiency", "CXCR4" ], "content": "Introduction\n\nToxoplasma gondii is an obligate intracellular protozoan with a wide host range among vertebrates, including humans1. Domestic cats and their relatives, the definitive hosts of T. gondii, release large numbers of unsporulated oocysts in their feces, which are then ingested by secondary hosts. Major sources of infection include ingestion of contaminated water or undercooked meat and exposure to other materials contaminated with cat feces, although transmission can also occur by transplantation, blood transfusion and laboratory accidents. Human infection has been estimated to occur in ~30% of individuals worldwide based on seroprevalence studies but usually results in lifelong subclinical infection in immunocompetent individuals. Chronic infection most commonly manifests as tissue cysts in skeletal muscle, myocardium, brain and eye. Acute toxoplasmosis in immunocompetent individuals may present as a mononucleosis-like syndrome. In the setting of acquired immunodeficiency, toxoplasmosis may occur as a result of primary T. gondii acquisition or reactivation of latent infection and may present as systemic illness or as a localized infection. Central nervous system toxoplasmosis is a particular problem in AIDS patients and is an AIDS-defining condition. Ocular toxoplasmosis can also occur in AIDS patients and may even be the presenting manifestation2. Vertical transmission of T. gondii is ~40% for women who become infected during pregnancy and may cause severe congenital developmental defects involving the brain, eye and other organs. In the eye, T. gondii may cause a progressive and recurring necrotizing retinochoroiditis and is the most common cause of infectious uveitis worldwide. Optic neuritis is a less frequent presentation that is usually associated with worse visual outcome. Ocular toxoplasmosis occurs in patients with acquired immunodeficiency but has not previously been reported in patients with primary immunodeficiency disorders.\n\nWarts-Hypogammaglobulinemia-Infections-Myelokathexis (WHIM) syndrome is a rare primary immunodeficiency disorder caused by autosomal dominant gain-of-function mutations in the chemokine receptor CXCR43. The primary manifestations of WHIM syndrome are cutaneous and mucosal warts, hypogammaglobulinemia, recurrent non-invasive infections, which are usually bacterial in origin and typically affect the oto-sino-pulmonary tract and skin, and myelokathexis, a term coined for retention of mature neutrophils in bone marrow. Opportunistic and life-threatening infections in WHIM syndrome patients are rare, and significant protozoan infection has not been previously reported.\n\n\nClinical course and management\n\nA 14-year-old Hispanic male with WHIM syndrome (mutation CXCR4R334X) from El Salvador presented with painless sudden onset right eye blindness of at least one week’s duration. There was no history of blunt or chemical trauma to the eye, recent bacterial or viral illness, or change in medication, and he reported no eye pain, periorbital swelling, eye discharge, fever, rash or headache. The past medical history was significant for Tetralogy of Fallot which had been repaired surgically. Neutropenia was diagnosed as a neonate, resulting in recurrent upper and lower respiratory tract infections complicated by bronchiectasis and tympanic membrane perforation. He received filgrastim (G-CSF, Amgen) from ages 1–3 but this was discontinued due to bone pain. At age 13, he developed dengue fever and three successive episodes of pneumonia, prompting evaluation for primary immunodeficiency. Panleukopenia was documented (ANC 90, AMC 50, ALC 1070, platelets 122,000, CD4+ T cells 365, CD19+ B cells 11 [all per microliter]). Vision was normal. After obtaining informed consent on an NIH IRB-approved protocol for immunodeficiency screening (05-I-0213), genetic testing of a blood sample identified heterozygous CXCR4R334X, the most common WHIM syndrome genotype. Three months later the patient experienced complete vision loss in the right eye but was otherwise asymptomatic.\n\nThe patient appeared well-developed but underweight (BMI = 14.5) with mild scoliosis. Splenomegaly was noted. Classic features of WHIM syndrome were present, including cutaneous warts, hypogammaglobulinemia, and severe panleukopenia (Table 1). A bone marrow biopsy revealed myelokathexis with an elevated 4:1 myeloid:erythroid ratio. In the right eye, light perception was absent and there was an afferent pupillary defect. Dilated fundus examination showed a pale retina with widespread white subretinal infiltrates with a necrotizing appearance in some areas, patches of subretinal fibrosis, mild vitritis and a fibrotic band reaching the optical nerve head and a pale optic nerve (Figure 1). Spectral domain optical coherence tomography images showed variable hyper-reflective infiltration of the subretinal space throughout the macula without serous subretinal fluid, with disruption of normal lamination of the macula. B-scan ultrasound showed mild vitreous opacities with presence of a posterior hyaloid membrane still adherent to the optic disc, but separated from the retina in other areas posteriorly, with presence of a vitreoschisis cavity inferiorly, without any retinal detachment, and without any definite eye wall thickening or episcleral lucency. The left eye was normal. Cranial CT scan showed mild sinusitis. Filgrastim (1 mcg/kg/day) was given resulting in increased ANC and increased vitritis the next day, suggesting the possibility of an ongoing chronic infection. Peripheral blood cultures were negative. A vitreous biopsy showed mixed inflammatory cells, and PCR testing was positive for T. gondii in vitreous fluid but negative in serum and bone marrow. Serum IgG for T. gondii was 599 international units/ml. Tests for other viral (CMV, EBV, VZV, HSV, dengue), fungal (Histoplasma, Cryptococcus), bacterial (Bartonella, Rickettsia, Legionella, Mycobacterium) and parasitic (Leishmania, Toxocara) pathogens were negative. A diagnosis of advanced ocular toxoplasmosis with ongoing active lesions was made.\n\nWBC: white blood cells; RBC: red blood cells; ANC: absolute neutrophil count; ALC: absolute lymphocyte count; AMC: absolute monocyte count; NK: natural killer cells. Measured values for the cell counts are cells/µL, except as otherwise noted.\n\nThe lower right panel shows the optical coherence tomography findings at presentation for OD (top) and OS (bottom).\n\nThe patient was treated with oral pyrimethamine (75 mg loading dose then 25 mg/day), leucovorin (7.5 mg/day), and sulfadiazine (1500 mg 2x/day) for six weeks. After treatment, chorioretinal scarring appeared stable. Serum IgG for T. gondii declined to 222 IU/ml 32 months later. The macula was fibrotic and atrophic without signs of active exudative lesions over 4 years follow up, during which the patient has received daily prophylactic trimethoprim/sulfamethoxazole (800 mg/160 mg). The optic nerve is atrophic in the right eye and normal in the left. Light perception continues to be absent in the right eye but left eye vision has remained normal. He has successfully completed a Phase 3 double blinded clinical trial (ClinicalTrials.gov Identifier NCT02231879) comparing 14 months each of twice daily plerixafor (Sanofi) and filgrastim (Amgen) treatment and is currently receiving open label filgrastim (1 mcg/kg/day). The patient has had markedly improved growth, no significant bacterial infections and is fully active, competing at the national level in equestrian sports.\n\n\nDiscussion\n\nTo our knowledge, this is the first detailed report of localized ocular toxoplasmosis in a primary immunodeficiency disorder and the first report of a protozoan infection in WHIM syndrome. In addition, optic nerve involvement as seen in our patient is rare in ocular toxoplasmosis (<5%)4.\n\nAlthough symptomatic toxoplasmosis occurs frequently in the setting of acquired immunodeficiency, especially HIV infection, it is rarely associated with a primary immunodeficiency disorder. Disseminated toxoplasmosis has been reported in several patients with X-linked hyper-IgM (XLHI) syndrome5–9. Neutropenia was observed in two of these patients5,7. Two patients were receiving IVIg replacement therapy at the time toxoplasmosis was diagnosed,5,8 while two others were previously undiagnosed with XLHI and had untreated hypogammaglobulinemia7,9. Of note, one patient had been taking trimethoprim/sulfamethoxazole as prophylaxis for recurrent otitis media prior to the onset of symptomatic toxoplasmosis5. Disseminated T. gondii infection with ocular manifestations has been reported in a patient with ataxia telangiectasia10. This patient did not have lymphopenia and had received IVIg replacement therapy. In addition, fatal cerebral toxoplasmosis was reported in two patients with common variable immunodeficiency11,12.\n\nThus, hypogammaglobulinemia is a common feature of primary immunodeficiency disorders in which toxoplasmosis has been reported, suggesting the importance of antibody-mediated immunity for controlling T. gondii. Although our patient had low immunoglobulins, he developed a strong IgG response to T. gondii. The quality of the antibodies and the kinetics of the response are unknown but evidently were insufficient to initially control the pathogen. Cell-mediated immunity is also important in control of T. gondii infection, with critical complementary roles for monocytes, neutrophils, dendritic cells, plasma cells, and CD4+ and CD8+ T cells13. IFNγ and IL-12 are hallmarks of the Th1 response to T. gondii infection13. Neutrophils, activated monocytes, macrophages, and dendritic cells all produce IL-12, whereas IFNγ is produced by NK cells, neutrophils, and effector T cells in response to T. gondii invasion13.\n\nAn explanation for the paucity of symptomatic T. gondii infections among patients with primary immunodeficiency is lacking. Possible explanations include the frequent use of broad-spectrum antibiotics such as trimethoprim-sulfamethoxazole for patients with primary and acquired immunodeficiency disorders and environmental precautions taken to limit exposure to pathogens in general.\n\nWHIM syndrome is a complex, phenotypically heterogenous primary immunodeficiency disorder that frequently involves defects in steady state levels of leukocytes and antibody in the blood, as in our patient. Given that the patient has multiple immunologic abnormalities, it is not possible to attribute his susceptibility to T. gondii to any single one.\n\nIn summary, we describe in detail a very rare case of primary ocular toxoplasmosis in primary immunodeficiency disease and the first case of protozoan infection in WHIM syndrome. The precise immunologic abnormalities among the spectrum of abnormalities resulting from WHIM syndrome that predisposed the patient to such a devastating outcome of T. gondii infection is currently unknown.\n\n\nConsent\n\nWritten informed pediatric assent was obtained from the patient, and the patient’s mother provided parental written informed consent for the publication of the patient’s clinical details and any associated images.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.", "appendix": "Grant information\n\nThis work was supported with federal funds from the Division of Intramural Research of the National Institute of Allergy and Infectious Diseases, National Institutes of Health. This project has also been funded in part with federal funds from the National Cancer Institute, National Institutes of Health, under Contract No. HHSN261200800001E.\n\nThe content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nMaenz M, Schluter D, Liesenfeld O, et al.: Ocular toxoplasmosis past, present and new aspects of an old disease. Prog Retin Eye Res. 2014; 39: 77–106. PubMed Abstract \| Publisher Full Text\n\nPomeroy C, Noble R, McCormick M, et al.: Ocular toxoplasmosis as the presenting manifestation of human immunodeficiency virus infection. Clin Infect Dis. 1997; 24(4): 745–6. PubMed Abstract \| Publisher Full Text\n\nHeusinkveld LE, Yim E, Yang A, et al.: Pathogenesis, diagnosis and therapeutic strategies in WHIM syndrome immunodeficiency. Expert Opin Orphan Drugs. 2017; 5(10): 813–25. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEckert GU, Melamed J, Menegaz B: Optic nerve changes in ocular toxoplasmosis. Eye (Lond). 2007; 21(6): 746–51. PubMed Abstract \| Publisher Full Text\n\nLeiva LE, Junprasert J, Hollenbaugh D, et al.: Central nervous system toxoplasmosis with an increased proportion of circulating gamma delta T cells in a patient with hyper-IgM syndrome. J Clin Immunol. 1998; 18(4): 283–90. PubMed Abstract \| Publisher Full Text\n\nLevy J, Espanol-Boren T, Thomas C, et al.: Clinical spectrum of X-linked hyper-IgM syndrome. J Pediatr. 1997; 131(1 Pt 1): 47–54. PubMed Abstract \| Publisher Full Text\n\nLiu X, Zhou K, Yu D, et al.: A delayed diagnosis of X-linked hyper IgM syndrome complicated with toxoplasmic encephalitis in a child: A case report and literature review. Medicine (Baltimore). 2017; 96(49): e8989. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTsuge I, Matsuoka H, Nakagawa A, et al.: Necrotizing toxoplasmic encephalitis in a child with the X-linked hyper-IgM syndrome. Eur J Pediatr. 1998; 157(9): 735–7. PubMed Abstract \| Publisher Full Text\n\nYong PF, Post FA, Gilmour KC, et al.: Cerebral toxoplasmosis in a middle-aged man as first presentation of primary immunodeficiency due to a hypomorphic mutation in the CD40 ligand gene. J Clin Pathol. 2008; 61(11): 1220–2. PubMed Abstract \| Publisher Full Text\n\nGioia LV, Bonsall D, Moffett K, et al.: Bilateral maculopathy in a patient with ataxia telangiectasia. J AAPOS. 2016; 20(1): 85–8. PubMed Abstract \| Publisher Full Text\n\nHoltkamp M, Okuducu AF, Klingebiel R, et al.: Cerebral toxoplasmosis in a patient with common variable immunodeficiency. Neurology. 2004; 63(11): 2192–3. PubMed Abstract \| Publisher Full Text\n\nShachor J, Shneyour A, Radnay J, et al.: Toxoplasmosis in a patient with common variable immunodeficiency. Am J Med Sci. 1984; 287(3): 36–8. PubMed Abstract \| Publisher Full Text\n\nDupont CD, Christian DA, Hunter CA: Immune response and immunopathology during toxoplasmosis. Semin Immunopathol. 2012; 34(6): 793–813. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "42400", "date": "21 Jan 2019", "name": "Sarah Beaussant-Cohen", "expertise": [ "Reviewer Expertise Primary immune deficiency." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nQuestion 1: Is the background of the case’s history and progression described in sufficient detail?\nDavid H. McDermott et al report the first advanced ocular toxoplasmosis complication in WHIM syndrome. The case is interesting because the patient exemplifies widely known aspects of the disease, demonstrating classic features of WHIM syndrome, while at the same time manifests aspects of the syndrome which clinicians unfamiliar with the disease may not be aware of such as biological combined immune deficiency or a history of Tetralogy of Fallot (Raffaele Badolato, J Pediatr. 2012)1.\nImportantly, this case exemplifies that manifestations reported in the acronym “W.H.I.M” (Warts, Hypogammaglobulinemia, Infections and Myelokathexis) insufficiently describe the disease and may even be misleading. Indeed, while the patient clearly presents three of the manifestations emphasized in the acronym 1) Warts 2) a history of recurrent upper and lower respiratory tract Infections and 3) Myelokathexis, the patient does not show biologically defined hypogammaglobulinemia. However, while his IgG levels are in the normal range (according to Table 1), his clinical presentation with repeated respiratory and ENT infections are typical of patients with hypogammaglobulinemia. Furthermore, in the discussion section, the authors suggest that the patient appears to have a qualitative humoral defect. This case clearly demonstrates that the spectrum of WHIM syndrome manifestations range well-beyond the acronym. Importantly, the reader will appreciate that the patient’s labs are compatible with combined immunodeficiency: CD4+ T cells 365, CD19+ B cells 11 [per microliter] and very low naive CD4+ and CD8+ T cells.\n\nWe can conclude that the authors have given an accurate description of a typical WHIM patient (CXCR4R334X ) who presents not only well-known manifestations of the disease (neutropenia, warts) as well as lesser known (Tetralogy of Fallot, combined immune deficiency) hallmarks of the disease. This detailed report will greatly help clinicians better apprehend and recognize this complex syndrome in their own patient population.\n\nQuestion 2: Are enough details provided of any physical examination and diagnostic tests, treatment given and outcomes?\n\nThe authors describe in detail the challenging diagnosis of toxoplasmic optic neuritis. As a reader, we understand the medical thought-process which led to the diagnosis. Due to the neutropenia, the patient did not initially show strong clinical findings of ocular toxoplasmosis. However, Filgrastim administration unmasked vitritis which is a prominent feature of ocular toxoplasmosis. The diagnosis was then confirmed by PCR testing which was positive for T. gondii in vitreous fluid. The patient received standard management of toxoplasmosis-associated optic neuropathy as he was treated with oral pyrimethamine (75 mg loading dose then 25 mg/day), leucovorin (7.5 mg/day), and sulfadiazine (1500 mg 2x/day) for six weeks (Rim Kahloun et al, Eye Brain. 2015)2. Optic neuritis remains an unusual presentation of ocular Toxoplasmosis that is associated with worse visual outcome. This is clearly shown during the prolonged follow-up of the patient who sustains findings of fibrotic and atrophic macula without signs of active exudative lesions.\n\nQuestion 3: Is sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment?\n\nThe authors are well aware that rare monogenic diseases such as WHIM syndrome offer a unique window of insight into the role of the CXCR4 pathway in health and disease. In their discussion, the authors offer a brief explanation of the possible pathophysiology of the appearance of toxoplasmic optic neuritis in this WHIM patient. They suggest that he presents with both impaired humoral immunity and defective cell-mediated immunity. In particular, they address the question of a defective IFN-gamma and/or IL-12 pathway in this disease. As of today, there is not a sufficient amount of work in the literature to address this question properly, nevertheless it may have been interesting to cite previous studies such as the work of Laura Tassone et al (Blood 2010) 3which suggested that mature DCs from WHIM patients produce normal amounts of interleukin-12 (p70) compared with the cells derived from healthy donors or the work of Marinos Kallikourdis et al (blood 2013- figure 5c)4 in which she shows that IFN-gamma production is not different between CD4+ T cells from a healthy donor or a WHIM patient (G336X) activated by anti-CD3– and anti-CD28–coated beads. Clearly, the questions raised by the authors are pertinent and underline the importance of further studies to better characterize the defects in this pathway.\n\nQuestion 4: Is the case presented with sufficient detail to be useful for other practitioners?\n\nWHIM is often classified as a severe congenital neutropenia, which may be misleading to clinicians unfamiliar with the disease. Therefore, this case will be very useful to clinicians involved in the care of patients with WHIM syndrome. This report will remind clinicians that they should keep a high degree of suspicion at all times as it is possible that their WHIM patients may present with unusual infections classically seen in patients with severe T cell immunodeficiency.\n\nMinor Changes to be addressed:\nTable 1:\nlines for both NK and NK-T are duplicated.\n\nPatient does not have IgG hypogammaglobulinemia according to reference ranges in table 1. This should be stated in the text.\n\nIt could be useful to bold the values with are outside of the normal range in table 1.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [] }, { "id": "49539", "date": "20 Jun 2019", "name": "Françoise Bachelerie", "expertise": [ "Reviewer Expertise HOST/VIRUS interactions with a special interest for the role of chemokine receptors in host surveillance." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIs the background of the case’s history and progression described in sufficient detail? David McDermott and colleagues are reporting the first example of protozoan infection (Toxoplasma gondii) in the context of the WHIM syndrome. The patient displays the characteristic clinical features of the syndrome including Human Papillomavirus (HPV)-induced warts, Myelokathexis and severe panleukopenia as detailed in the manuscript. He harbors a well-known CXCR4 mutation (R334X) most likely inherited from one of his parents (although sporadic cases were reported) but this information is not provided. Along this line, whether parents or close relatives also present with serum IgG for T. gondii is likely but not provided. It is interesting to note that the patient developed dengue virus infection one year before being diagnosed for Toxoplasma gondii but was negative for serum IgG (witnessing a primary infection?).\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Diagnostic for Toxoplasma gondii infection and treatment given are clearly mentioned. It would be also interesting that the authors mention about the evolution of Toxoplasma gondii infection by PCR testing during the 4 years follow up in the course of the daily trimethoprim/sulfamethoxazole treatment. The authors mention that infection was detected in vitreous fluid but not in serum nor in bone marrow. Were they other sites tested that would help to better understand the protozoan infection pathogenesis (acquisition/reactivation)? The authors mention that the patient successfully completed phase 3 plerixafor and filgrastim-based clinical trial. It will be thus beneficial for a better understanding of the patient treatment and outcome, to indicate when the patient was enrolled into this trial with regard to the T. gondii 4 year-treatment and, if appropriate, whether any follow up of the infection and responses toward it (e.g. serum Ig) was performed in the course of this trial.\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? As emphasized by David McDermott and colleagues, protozoan infections are rare in primary immunodeficiency disorders and this is first example of such infection in the context of the WHIM syndrome. Authors provide interesting hypotheses that might account for the clinical manifestations of Toxoplasma gondii infection and pathogenic mechanisms. I would recommend them to add recent literature regarding the potential role of CXCR4 in the life cycle of other members of the phylum Apicomplexa (Plasmodium) assessed in rodent models1. For other parasites such as helminths (e.g. L. sigmodontis) analyzed in rodent models of infection, neutrophils (and their control by a CXCR4/CXCL12 dependent mechanism) were proposed to be critical elements of the host innate protective response2.\nIs the case presented with sufficient detail to be useful for other practitioners? The report of this singular case is of great interest for scientists and clinicians working in the field and certainly very useful for practitioners not well aware of the syndrome.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [] }, { "id": "49621", "date": "10 Jul 2019", "name": "Lynnette J. Mazur", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe Center for Disease Control notes that ‘parasitic infections are typically associated with poor and often marginalized communities in low-income countries. However, these infections are also present in the United States.’ Additionally, they have designated toxoplasmosis a neglected parasitic infection and a public health priority because of the little attention it has received1. Based on the number of people infected, the severity of illness, and the ability for prevention and treatment, health care providers need to take note of its increasing prevalence along with the common signs and symptoms. The case report by McDermott et al. provides this opportunity.\n\nThe report details a teen with a WHIM syndrome, a primary immunodeficiency disorder, who developed sudden painless right eye blindness due to ocular toxoplasmosis (OT) or toxoplasmic choroiditis2. However, it is unknown whether his infection was congenitally acquired or acquired later in life through contaminated food or water or possibly as a result of a blood transfusion during his Tetralogy of Fallot repair. The maternal prenatal history, birth history, and the timing of his immigration might yield some clues. There is evidence that persons with congenital infection have a higher incidence of macular involvement, as in this case, and therefore, have an increased risk of legal blindness3.\nAdditionally, testing for IgM and performing a quantitative IgG avidity test may help differentiate acute from chronic infection. The avidity test is based on the immovability antigen and antibody attachment where the IgG antibody binds weakly to the antigen during the initial stage but over weeks or months progressively increases45. Patients with positive IgM and/or low IgG avidity are considered as acutely acquired cases (usually defined as acquisition of infection in recent six months. Alternately, a negative IgM and a high IgG avidity index support a more chronic infection. A recent study on immune mediators in the aqueous humor showed that IL-9 and IL-13 were significantly higher in primary OT than in recurrent OT6.\nCases of isolated OT commonly results from reactivation of untreated congenital infection which can occur years after birth. However, it also occurs in a small percentage of people with acquired infection, especially in patients that are immunocompromised. This case heightens the awareness of the signs and symptoms, and consequences of toxoplasmosis infection for healthcare providers. It serves as an opportunity to educate patients, especially in international persons and immunocompromised individuals on common exposures, preventive measures, and treatment.\n\nIs the background of the case’s history and progression described in sufficient detail? Partly\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Partly\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-2
https://f1000research.com/articles/8-1115/v1	17 Jul 19	{ "type": "Research Article", "title": "Wear assessment of current aesthetic crowns compared with human enamel after two finishing procedures: In vitro study", "authors": [ "Ahmed M. Sha'aban", "Gihan A. El Naggar", "Rasha Nabil", "Mohamed A. Rashad", "Yara S. Attia", "Gihan A. El Naggar", "Rasha Nabil", "Mohamed A. Rashad", "Yara S. Attia" ], "abstract": "Background: Surfaces of ceramic crowns are modified several times before being exposed to wear in the oral cavity. Grinding and different finishing procedures may be associated with teeth wear due to increased surface roughness. Limited data is available with regard to the effect of polishing procedures on the surface roughness and wear behavior of ceramic crowns. This study was conducted to assess the influence of polished and polished-ground-repolished surface finish on the roughness and wear performance of three ceramic crowns. Methods: 36 natural 1st molar teeth were prepared using a CNC milling machine and classified into three groups (n=12/group): zirconia, E-max and hybrid ceramic (VITA ENAMIC) crowns. Each group was classified into two subgroups (n=6/subgroup): polished and polished-ground-repolished crowns. Natural molar teeth served as an unrestored control group (n=6). All samples were loaded into a chewing simulator for 100,000 cycles and subjected to 600 thermo-cycles in temperature changes to simulate changes in intraoral temperature. Natural maxillary 1st premolar teeth were collected and only buccal half (cusp) of sectioned tooth was used as antagonists. A profilometer was used to detect the roughness before and after masticatory cycles. The occlusal surface was analysed using a scanning electron microscope (SEM). Results: The E-max crown group had the highest mean surface roughness value (0.267µm) followed by VITA ENAMIC crown group (0.266 µm), while the lowest mean surface roughness value recorded for zirconia crown (0.257µm). The difference between these means was not significant. The polished-ground–repolished group had a higher mean surface roughness (0.266 µm) compared with the polished group (0.260 µm), which was not significantly different. Conclusions: All tested ceramic crowns showed surface roughness with values within acceptable clinical parameters (~0.2 µm). Additionally, intraoral polishing procedures could be considered a reliable technique for smoothing of zircona, E.max and VITA ENAMIC crowns after occlusal adjustment.", "keywords": [ "Teeth wear", "Full ceramic crowns", "CAD/CAM", "surface roughness", "Enamel" ], "content": "Introduction\n\nWear of teeth is considered one of the most important issues when choosing a dental ceramic1,2. The wear may affect the thickness of the restoration and the natural enamel, which impairs the esthetic appearance and limits tooth survival rates3,4. Recent studies suggest that surface roughness highly influences o tooth wear5. A smooth ceramic surface enhances the function and aesthetics of teeth, and prevents gingival inflammation and the accumulation of stains and plaque on the surface of a restoration6, while a rough surface may abrade natural teeth, which compromises the appearance of the restoration7.\n\nThe repeated incidence of veneer chipping with bilayered restorations has prompted the development of monolithic restorations fabricated from high strength ceramics8,9. The present research was performed to compare the influence of polished and repolished procedures on surface roughness of the tooth and what impact on the wear rate of monolithic crowns this would have.\n\n\nMethods\n\nThe materials tested in the study are listed in Table 1.\n\nIn total, 42 intact mandibular 1st molar teeth with three roots were collected from patients (age range 40–50 years) from the Dental Surgery Clinic at Faculty of Dentistry, Cairo University, as part of routine periodontal treatment. Patients provided written informed consent for their extracted teeth to be used for subsequent research purposes.\n\nAfter removal, each tooth was sterilized in 0.5% sodium hypochlorite (ADWIC Egypt) and stored in saline solution. All teeth were mounted in epoxy resin (IN2 INFUSION; Easy Composites, USA) blocks using a plastic ring (25 × 35 mm). To simulate the function of the periodontium, the root was covered with thin layer of condensation silicone (Zetaplus; ZhermackSpA, Italy)10. A specially designed centralizing device was fabricated to allow accurate centralization of the tooth in the plastic ring during construction of the epoxy resin block (Figure 1).\n\nTeeth were prepared in a standardized manner using a CNC milling machine (CNC Premium 4820-i-mes, Germany). The external surface of the tooth was reduced to remove the height of contour creating a 1mm deep chamfer finish line with a 10° total convergence angles. The height of tooth was reduced 2 mm with a flat occlusal table (Figure 2 and Figure 3).\n\nImage created by authors using Autodesk Maya Software 2016.\n\nFabrication of the crowns. In order to standardize the shape and dimensions of all samples, CAD/CAM technology (CEREC in-Lab SW 15.1) was used according to the instructions of the manufacturer for different types of ceramic crowns. To obtain a 3D image, the prepared tooth was sprayed using an optical reflection medium (Optispray; Sirona Dental, Bensheim, Germany) and scanned using the CEREC Omnicam (Sirona Dental). After the scanned image was displayed on the computer screen, all scanning parameters (length: 8.05 mm; width: 7.19 mm; height: 6.34mm) of each tooth were restored in computer software (CEREC in-Lab software version 15.1, Sirona Dental).\n\nSintering of the zirconia ceramic crowns. Firing of zirconia crowns in a sintering furnace (TABEO Furnace; MIHM-VOGT, Germany) was carried out at approximately 9 hours, reaching around 1500°C within 4 hours then decreased to room temperature within 3 hours. All zirconia crowns were checked over their corresponding teeth for seating and inspected to detect eventual defects generated by sintering.\n\nCrystallization of E.max CAD ceramic crowns. Programat P510 furnace (Ivoclar Vivadent, Schan, Liechtenstein) was used. The starting temperature was 403°C and increased to 820°C, which was held for 10 minutes; then increased to 840°C, which was held for 7 minutes. After finishing of the firing cycle, crowns were removed and allowed to cool at room temperature. All crowns were checked over their corresponding teeth for seating and inspected to detect eventual defects generated by crystallization.\n\nZirconia crowns. Crowns were polished with a special polishing kit (Dedeco, Hi-glaze polishing Bur, USA) according to the manufacturer’s guidelines in a 3-step procedure at 10.000 rpm with a laboratory micromotor (Forte 100 Dental Micro Motor Unit Handpiece; Saeshin, Korea) under water cooling for 30 seconds in one direction with a constant light pressure by the same operator11–13. At the final step, Pearl surface Z was added to the polisher (fine grit size) to polish the entire crown surface14 without water to obtain a high luster gloss surface.\n\nE-max CAD and VITA ENAMIC crowns. VITA ENAMIC crowns were prepolished by the disc of the medium grit size at 7.000 rpm hand piece under dry condition for 1 minute. Then, the polisher of extra-fine grit size was used at 5000 rpm. Wet polishing of E-max CAD crown was performed using VITA ENAMIC technical kit similar to the method used for VITA ENAMIC crowns.\n\nZirconia crowns. Zirconia crowns were treated with 110 μm alumina particles using an airborne-particle-abrasion device (UMG Dental Lab Sandblaster; Tangshan UMG Medical Instrument, China) for 20 second15.\n\nE-max CAD and VITA ENAMIC crowns. The internal surfaces of crowns were treated with 5-% hydro-fluoric acid gel. Then, silane coupling agent (BISCO, Schaumburg, USA) was applied.\n\n35% phosphoric acid (3M ESPE) was placed on the teeth for 15 seconds and then rinsed for 10 sec. A thin layer of dentin bonding agent (3M ESPE, Paul, USA) was applied and polymerized (Elipar II) for 10 sec. Rely x ultimate resin cement was applied onto the internal surface of the crown which was then placed in position with gentle finger pressure. After that, the crowns were polymerized for 40 sec and then stored in H2O for 24 hrs16.\n\nHalf of each group (zirconia, E-max CAD and VITA ENAMIC groups (n=18)) were ground at the functional cusp (mesiobuccal and distobuccal cusp) with a fine diamond bur under water spray for 15 sec in sweeping movement by a single operator.\n\nZirconia crowns. Crowns were repolished again by a single operator, using Kenda Zircovis polishing system with a low speed handpiece under standardized conditions (permanent water cooling, 30 second, 10,000 rpm and a constant light pressure) in sweeping motion12,17.\n\nE-max CAD and VITA ENAMIC crowns. E-max CAD and VITA ENAMIC crowns were repolished again using Kenda Nobilis polishing system under standardized conditions (permanent water cooling, 30 second, 10,000 rpm and a constant light pressure) as suggested by the manufacturer (Figure 4).\n\nSelection of the abrader (antagonist) specimen. In total, 42 sound and freshly extracted maxillary 1st premolar teeth were collected from patients (age range, 15–20 years) from the Dental Orthodontic Clinic at Faculty of Dentistry, Cairo University as a part of orthodontic treatment. After removal, the teeth were sterilized in 0.5% sodium hypochlorite (ADWIC Egypt) and transferred to distilled saline solution.\n\nPreparation of human enamel antagonist specimens. Each tooth was sectioned vertically (bucco-palatally) into two halves using a diamond cutting disk (Single and Double sided diamond discs, NTI Inter flex, Kerr, Italia) attached to a low speed hand piece (Dental Latch Low Speed Contra-Angle Handpiece, NSK, Japan) with coolent. Only buccal half (cusp) of sectioned 1st premolar tooth was used as antagonists4. Jackob’s chuck holder was designed to mount the coronal portion of sectioned maxillary 1st premolar tooth (Figure 5).\n\nWear testing. Samples were loaded into a chewing simulator (ROBOTA chewing simulator; ROBOTA Model ACH-09075DC-T, LTD., Germany) for 100,000 cycles and subjected to 600 thermo-cycles (laboratory method of exposing ceramic crowns to temperature ranges similar to those occurring in the oral cavity) between 5⍜C and 55 ⍜C for 60 second each which simulate changes in the intraoral temperature, maintained by a thermo-statically controlled liquid circulator (Model ACH-09075DC-T, AD-TECH TECHNOLOGY CO., LTD., Germany) (Figure 6).\n\nThe samples were embedded in Teflon housing in the lower sample holder which was filled with water and kept the samples wet during the testing and removed the debris from the surface. The buccal half (cusp) of sectioned 1st premolar tooth antagonist was positioned on the ceramic crown to achieve wear contact with lateral movement of 2 mm and loaded with 49 Newton, which is equivalent to a normal chewing force in the oral cavity (Figure 7). The test was repeated 100,000 times to clinically simulate a six month chewing condition9,18.\n\nImage created by authors using Autodesk Maya Software 2016.\n\nThe 3D surface of the crowns were analyzed before and after the wear test using an USB digital microscope (Aven 26700-300 ZipScope USB Digital microscope, WITec, China) with a magnification of 120 X for roughness measurement19. The surface roughness expressed in micron (μm) was calculated using a digital image analysis system (WSxM software, Version 5 develop 4.1, Nanotec, Electronica, SL, Spain) by superimposing the 3D surfaces before and after the wear test (Figure 8–Figure 10). The images were recorded with a resolution of 1280 × 1024 pixels per image.\n\nPolished with (a) Dedco kit; (b) Kenda kit.\n\nPolished with (a) Technical kit; (b) Kenda kit.\n\nPolished with (a) Technical kit; (b) Kenda kit.\n\nOne sample of each subgroup was scanned using a scanning electron microscope (HITACHI, SU8200 series, Tokyo, Japan) at X500 magnification.\n\nTwo-factorial ANOVA was done to detect effect of two different types of polishing on the surface roughness of various ceramic crown (zirconia, E-max CAD, VITA ENAMIC). GraphPad Instat statistics software (Graph Pad InStat, Version 3, U-S-A) for Windows was used for statistical analysis. P values ≤ 0.05 are considered to be statistically significant in all tests.\n\n\nResults\n\nSurface roughness between all groups (before chewing simulation) was not statistically significant (Table 2).\n\nns: non-significant (p>0.05). (Control, n=6; zirconia/E.max/VITA ENAMIC, n=6/group)\n\nSurface roughness after chewing simulation revealed that; E.max re-polished crown had the highest surface roughness (0.2725 µm), while the lowest surface roughness value was recorded for the zirconia re-polished group (0.2570 µm). The different between surface roughness for all three crowns after chewing simulation was not statistically significant (P=0.666; Table 3).\n\nns: non-significant (p>0.05)\n\nSEM images for zirconia crowns, both polished and repolished, showed the smoothest surface compared with the other two crowns. For all polished-ground-repolished samples, surface changes, such as pitting, ridges and deep grooves, were observed on the ceramic (Figure 11–Figure 13).\n\n(a) Polished sample, some remnants of the polishing paste, narrow voids and smooth scar (black arrows) observed; (b) ground sample, marked major void defects (white arrow) observed; (c) repolished sample, smooth shallow paralleled striation observed.\n\n(a) Polished sample, deep grooves observed (white arrows); (b) ground sample, formation of ridges and deep grinding grooves observed (white arrows); (c) repolished sample, small voids (white arrows) and direction of polishing (black arrow) observed.\n\n(a) Polished sample, wide smooth dotted area observed; (b) ground sample, formation of rough parallel furrows observed; (c) repolished sample, smoothed wide spread pitted area observed.\n\n\nDiscussion\n\nThe advancement of dentistry has made it possible to construct full monolithic crowns with improved wear resistance. Intraoral adjustments of ceramic crowns may result in surface roughness; therefore, technical and intraoral polishing is a necessary procedure20. Numerous studies have compared different finishing procedures;21 however, there is little research comparing different finishing procedures and the resulting wear behavior.\n\nIn the present study, natural teeth were used as they have characteristics closer to the clinical condition22. The teeth were embedded in epoxy blocks simulating alveolar bone23, and a specially designed centralizing device was used, which was an important step for accurate centralization of the tooth in the epoxy blocks24. A constant seating force by the fingertip of a single operator was applied before curing to simulate the clinical cementation steps carried out by operators in clinical practice25,26.\n\nThe polishing procedures with dental laboratory polishing kits were chosen to simulate the surface conditions after milling. In the current study, samples were roughened with a fine diamond instrument, which ensures more uniform grind on the surface27. The same operator with a light pressure was used throughout the study to polish all samples in order to standardize the procedures11–13.\n\nThe samples were subjected to 600 thermo-cycles between 5ºC and 55ºC to simulate changes in the intraoral temperature. This thermal-cycling caused additional ageing of the samples28. The chewing force of the wear test device was 49 Newton with a frequency of ~1–1.6 Hz which simulated average chewing load in the oral cavity and equals the normal occlusal force reported in previous studies12,29.\n\nRoughness values obtained after repolishing procedures were within the clinically acceptable value as reported by previous studies12,30. It could be assumed that the tested ceramic repolished surfaces produced only minimal wear to the opposing enamel (neglected wear). This also may indicate the efficiency of the polishing procedures carried out after grinding in smoothing the resulting roughness, which may explain the non-significant difference between the effect of the two surface finishings tested regardless of the characteristic differences of the three ceramic materials tested under chewing simulation. The results of the SEM were in agreement with previous studies11,31 as they concluded that finishing procedures might remove some of the defects on the ceramic surface.\n\nWhen the effect of crown material is considered, the results revealed that the highest surface roughness value was in E.max ceramic crowns (0.267), then VITA ENAMIC ceramic crowns (0.266), while the lowest roughness was recorded for zirconia crown (0.257), but these differences were not significant. Rashid32 previously said that despite the variation in base line crystalline structures between these ceramics, smooth surface can be produced for different ceramics with the same polishing steps. Therefore, the present results were not a surprising result, as the same kits were used in the current study.\n\n\nConclusion\n\nAll tested crowns showed surface roughness within acceptable clinical parameters when subjected to a different surface finish and six-month chewing simulation.\n\nNo change in the wear behavior was encountered with the three tested ceramic materials with the applied surface finishing procedures.\n\nIntraoral polishing procedures could be considered a reliable technique for smoothing of the zircona, E.max and VITA ENAMIC crowns after occlusal adjustment.\n\n\nData availability\n\nOpen Science Framework: Wear Assessment of Current Esthetic Crowns Against Human Enamel After Two Finishing Procedures: In vitro study, https://doi.org/10.17605/OSF.IO/MYGWD33.\n\nThis project contains the following underlying data:\n\nZip file containing images of digital microscope for each crown, before and after six month chewing simulation.\n\nResults.xls: raw surface roughness values for each crown and treatment, before and after six month chewing simulation.\n\nSEM raw, unedited images for those shown in Figure 11–Figure 13.\n\n3D topography files for the images shown Figure 8–Figure 10.\n\nCEREC software images.\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nThe authors declare that no grants were involved in supporting this work.\n\n\nReferences\n\nAhmadzadeh A, Hashemi Ashtiani AR, Epakchi S: Comparison of the Effect of Feldspathic Porcelain and Zirconia on Natural Tooth Wear. J Islam Dent Ass Iran. 2014; 26(2): 89–95. Reference Source\n\nMaglad AS, Wassell RW, Barclay SC, et al.: Risk management in clinical practice. part 3. Crowns and bridges. Braz Dent J. 2010; 209(3): 115–22. PubMed Abstract \| Publisher Full Text\n\nPassos SP, Torrealba Y, Major P, et al.: In vitro wear behavior of zirconia opposing enamel: A systematic review. J Prosthodont. 2014; 23(8): 593–601. PubMed Abstract \| Publisher Full Text\n\nHeintze SD, Cavalleri A, Forjanic M, et al.: Wear of ceramic and antagonist--a systematic evaluation of influencing factors in vitro. Dent Mater. 2008; 24(4): 433–49. PubMed Abstract \| Publisher Full Text\n\nSaiki O, Koizumi H, Akazawa N, et al.: wear characteristics of polished and glazed lithium disilicate ceramics opposed to three ceramic materials. J Oral Sci. 2016; 58(1): 117–23. PubMed Abstract \| Publisher Full Text\n\nSilva TM, Salvia AC, Carvalho RF, et al.: Effects of Different Polishing Protocols on Lithium Disilicate Ceramics. Braz Dental J. 2015; 26(5): 478–83. PubMed Abstract \| Publisher Full Text\n\nBoaventura J, Nishida R, Elossais AA: Effect finishing and polishing procedures on the surface roughness of IPS Empress 2 ceramic. Acta Odontol Scand. 2013; 71(3–4): 438–43. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSilva LH, Lima E, Stephanie S, et al.: Dental ceramics: a review of new materials and processing methods. Braz Oral Res. 2017; 31(suppl 1): e58. PubMed Abstract \| Publisher Full Text\n\nNosti J: Current Guide to All Ceramic Restoration. Dent Mater. 2012; 28(4): 106–112. Reference Source\n\nMarchionatti AM, Wandscher VF, Broch J, et al.: Influene of peridontal ligament simulation on bond strength and fracture resistance of roots restored with fiber posts. J App Oral Sci. 2014; 22(5): 450–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBassir M, Babasafari M, Rezvani MB, et al.: Effect of coarse grinding, overglazing, and 2 polishing systems on the flexural strength, surface roughness, and phase transformation of yttrium-stabilized tetragonal zirconia. J Prosthet Dent. 2017; 118(5): 658–665. PubMed Abstract \| Publisher Full Text\n\nPreis V, Weiser F, Handel G, et al.: Wear performance of monolithic dental ceramics with different surface treatments. Quintessence Int. 2013; 44(5): 393–405. PubMed Abstract \| Publisher Full Text\n\nPereira GK, Amaral M, Simoneti R, et al.: Effect of grinding with diamond-disc and -bur on the mechanical behavior of a Y-TZP ceramic. J Mech Behav Biomed Mat. 2014; 37: 133–40. PubMed Abstract \| Publisher Full Text\n\nChevalier J, Grenmillard L, Virkar AV, et al.: The Tetragonal-Monoclinic Transformation in Zirconia: Lessons Learned and Future Trends. J Am Ceram Soc. 2009; 92(9): 1901–1920. Publisher Full Text\n\nCurtis AR, Wright AJ, Fleming GJ: The influence of surface modification techniques on the performance of a Y-TZP dental ceramic. J Dent. 2006; 34(3): 195–206. PubMed Abstract \| Publisher Full Text\n\nPatel P, Thummar M, Shah D, et al.: Comparing the Effect of a Resin Based Sealer on Crown Retention for Three Types of Cements: An In Vitro Study. J Indian Prosthodont Soc. 2013; 13(3): 308–14. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBassir M, Bagher MR, Golzari H: Effect of Two Polishing Systems on Surface Roughness, Topography, and Flexural Strength of a Monolithic Lithium Disilicate Ceramic. J Prosthodont. 2017; 1–9.\n\nBasunbul G, Nathanson D: Human enamel wear against four dental ceramic in vitro. 2011. Reference Source\n\nKakaboura A, Fragouli M, Rahiotis C, et al.: Evaluation of surface characteristics of dental composites using profilometry, scanning electron, atomic force microscopy and gloss-meter. J Mater Sci. 2007; 18(1): 155–163. PubMed Abstract \| Publisher Full Text\n\nPreis V, Schmalzbauer M, Bougeard D, et al.: Surface properties of monolithic zirconia after dental adjustment treatments and in vitro wear simulation. J Dent. 2015; 43(1): 133–139. PubMed Abstract \| Publisher Full Text\n\nSabrah AH, Cook NB, Luangruangrong P, et al.: Full-contour Y-TZP ceramic surface roughness effect on synthetic hydroxyapatite wear. J Dent Mater. 2013; 29(6): 666–673. PubMed Abstract \| Publisher Full Text\n\nHeintze SD: How to qualify and validate wear simulation devices and methods. Dent Mater. 2006; 22(8): 712–734. PubMed Abstract \| Publisher Full Text\n\nBerkovitz B, Holland G, Moxham B: Oral anatomy, histology and embryology. 4th ed. St. Louis: Elsevier; 2009; 140–175. Reference Source\n\nZhi-Yue L, Yu-Xing Z: Effects of post-core design and ferrule on fracture resistance of endodontically treated maxillary central incisors. J Prosthet Dent. 2003; 89(4): 368–373. PubMed Abstract \| Publisher Full Text\n\nPeerzada F, Yiu CK, Hiraishi N, et al.: Effect of surface preparation on bond strength of resin luting cements to dentin. Oper Dent J. 2010; 35(6): 624–633. PubMed Abstract \| Publisher Full Text\n\nSreeramulu B, Suman P, Ajay P: A comparison between different luting cements on the retention of complete cast crowns - an in vitro study. Int J Health Biomed Res. 2015; 3(4): 29–35. Reference Source\n\nAldegheishem A, Alfaer A, Brezavšček M, et al.: Wear behavior of zirconia substrates against different antagonist materials. Int J Esthet Dent. 2015; 10(3): 468–485. PubMed Abstract\n\nMörmann WH, Stawarczykb B, Endera A, et al.: Wear characteristics of current aesthetic dental restorative CAD/CAM materials: two-body wear, gloss retention, roughness and Martens hardness. J Mech Behav Biomed Mater. 2013; 20: 113–125. PubMed Abstract \| Publisher Full Text\n\nHmaidouch R, Weigl P: Tooth wear against ceramic crowns in posterior region: a systematic literature review. Int J Oral Sci. 2013; 5(4): 183–190. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPark JH, Park S, Lee K, et al.: Antagonist wear of three CAD/CAM anatomic contour zirconia ceramics. J Prost Dent. 2014; 111(1): 20–29. PubMed Abstract \| Publisher Full Text\n\nHmaidouch R, Müller WD, Lauer HC, et al.: Surface roughness of zirconia for full-contour crowns after clinically simulated grinding and polishing. Int J Oral Sci. 2014; 6(4): 241–246. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRashid H: The effect of surface roughness on ceramics used in dentistry: A review of literature. Eur J Dent. 2014; 8(4): 571–579. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSha'aban AM: Wear Assessment of Current Esthetic Crowns Against Human Enamel After Two Finishing Procedures: In Vitro Study. OSF. 2019. http://www.doi.org/10.17605/OSF.IO/MYGWD" }	[ { "id": "70504", "date": "18 Sep 2020", "name": "Hao Yu", "expertise": [ "Reviewer Expertise Fixed Prosthodontics", "Material Sciences" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors investigated an area of clinical relevance. In general, the study was well designed and straighforward. The reviewer has the following comments that the authors may consider to improve the article:\nThe Authors are suggested to present the data in the results section with emphasis on the statistical analysis, but not the mean values.\n\nFor the grouping methods, why not add one group of polished-ground crowns as positive control?\n\nPlease state the rationale of adopting the current chewing cycles and thermocycling.\n\nIn the Abstract section, it was written \"A profilometer was used....\", which is controversial to the description in the M&M section.\n\nPlease add essential description of the polishing systems used in the present study. The reviewer is confused with the polishing systems used in different groups. The authors used \"Kenda\" kits for all the repolishing? if so, it seems unreasonable.\n\nWhat is the accuracy and detection limit of the USB digital microscope?\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/8-1115
https://f1000research.com/articles/8-377/v1	04 Apr 19	{ "type": "Software Tool Article", "title": "CGAT-core: a python framework for building scalable, reproducible computational biology workflows", "authors": [ "Adam P. Cribbs", "Sebastian Luna-Valero", "Charlotte George", "Ian M. Sudbery", "Antonio J. Berlanga-Taylor", "Stephen N. Sansom", "Tom Smith", "Nicholas E. Ilott", "Jethro Johnson", "Jakub Scaber", "Katherine Brown", "David Sims", "Andreas Heger", "Sebastian Luna-Valero", "Charlotte George", "Ian M. Sudbery", "Antonio J. Berlanga-Taylor", "Stephen N. Sansom", "Tom Smith", "Nicholas E. Ilott", "Jethro Johnson", "Jakub Scaber", "Katherine Brown" ], "abstract": "In the genomics era computational biologists regularly need to process, analyse and integrate large and complex biomedical datasets. Analysis inevitably involves multiple dependent steps, resulting in complex pipelines or workflows, often with several branches. Large data volumes mean that processing needs to be quick and efficient and scientific rigour requires that analysis be consistent and fully reproducible. We have developed CGAT-core, a python package for the rapid construction of complex computational workflows. CGAT-core seamlessly handles parallelisation across high performance computing clusters, integration of Conda environments, full parameterisation, database integration and logging. To illustrate our workflow framework, we present a pipeline for the analysis of RNAseq data using pseudo-alignment.", "keywords": [ "workflow", "pipeline", "python", "genomics" ], "content": "Introduction\n\nGenomic technologies have given researchers the ability to produce large amounts of data at relatively low cost. Bioinformatic analyses typically involve passing data through a series of manipulations and transformations, called a pipeline or workflow. The need for tools to manage workflows is well established, with a wide range of options available from graphical user interfaces such as Galaxy1 and Taverna2, aimed at non-programmers, to Snakemake, Nextflow, Toil, CGAT-Ruffus and others3–10 developed with computational biologists in mind. These tools differ in their portability, scalability, parameter handling, extensibility, and ease of use. In a recent survey11, the tool rated highest for ease of pipeline development was CGAT-Ruffus12, a Python package that wraps pipeline steps in discrete Python functions, called ‘tasks’. It uses Python decorators to track the dependencies between tasks, ensuring that dependent tasks are completed in the correct order and independent tasks can be run in parallel. If a pipeline is interrupted before completion, or new input files are added, only data sets that are missing or out-of-date are re-run. CGAT-Ruffus implements a wide range of decorators that allow complex operations on input files including: conversion of a single input file to a single output file; splitting of a file into multiple files (and vice versa) and conditional merging of multiple input files into a smaller number of outputs. More advanced options include combining combinations or permutations of input files and conditional execution based on input parameters. Use of decorators means that CGAT-Ruffus pipelines are native Python scripts, rather than the domain specific languages (DSLs) used in other workflow tools. A key advantage of this, in addition to python being an already widely understood language in computational biology, is that individual steps can use arbitrary python code, both in how they are linked together and in the actual processing task.\n\nHere, we introduce Computational Genomics Analysis Toolkit (CGAT)-core13, an open-source python library that extends the functionality of CGAT-Ruffus by adding cluster interaction, parameterisation, logging, database interaction and Conda environment switching.\n\n\nMethods\n\nCGAT-core13 extends the functionality of CGAT-Ruffus by providing a common interface to control distributed resource management systems using Distributed Resource Management Application API (DRMAA),. Currently, we support interaction with Sun Grid Engine, Slurm and PBS-pro/Torque. The execution engine enables tasks to be run locally or on a high-performance computing cluster and supports cluster distribution of both command line scripts (cgatcore.run) and python functions (cgatcore.cluster). System resources (number of cores to use, amount of RAM to allocate) can be set on a per-pipeline, per-task, or per task-instance basis, even allowing allocation to be based in variables, for example input file size.\n\nThe parameter management component encourages the separation of workflow/tool configuration from implementation to build re-usable workflows. Algorithm parameters are collected in a single human-readable yaml configuration file. Thus, parameters can be set specifically for each dataset, without the need to modify the code. For example, sequencing data can be aligned to a different reference genome, by simply changing the path to the genome index in the yaml file. Both pipeline-wide and job-local parameters are automagically substituted into command line statements at execution-time.\n\nTo assist with reproducibility, record keeping and error handling CGAT-core provides multi-level logging during workflow execution, recording full details of runtime parameters, environment configuration and tracking job submissions. Additionally, CGAT-core provides a simple, lightweight interface for interacting with relational databases such as SQLite (cgatcore.database), facilitating loading of analysis results at any step of the workflow, including combining output from parallel steps in single wide- or long-format tables.\n\nCGAT-core can load a different Conda environment for each step of the analysis, enabling the use of tools with conflicting software requirements. Furthermore, providing Conda environment files alongside pipeline scripts ensures that analyses can be fully reproduced.\n\nCGAT-core workflows are Python scripts, and as such are stand-alone command line utilities that do not require the installation of a dedicated service. In order to reproducibly execute our workflows, we provide utility functions for argument parsing, logging and record keeping within scripts (cgatcore.experiment). Workflows are started, inspected and configured through the command line interface. Therefore, workflows become just another tool and can be re-used within other workflows. Furthermore, workflows can leverage the full power of Python, making them completely extensible and flexible.\n\n\nImplementation\n\nCGAT-core is implemented in Python 3 and installable via Conda and PyPI with minimal dependencies. We have successfully deployed and tested the code on OSX, Red Hat and Ubuntu. We have made CGAT-core and associated repositories open-source under the MIT licence, allowing full and free use for both commercial and non-commercial purposes. Our software is fully documented (https://pypi.org), version controlled and has extensive testing using continuous integration (https://travis-ci.org/cgat-developers.) We welcome community participation in code development and issue reporting through GitHub.\n\n\nUse case\n\nTo illustrate a simple use case of CGAT-core, we have built an example RNAseq analysis pipeline, which performs read counting using Kallisto14 and differential expression using DESeq215. This workflow and Conda environment are contained within our CGAT-showcase repository (https://github.com/cgat-developers/cgat-showcase). The workflow highlights how simple pipelines can be constructed using CGAT-core, demonstrating how the pipeline can be configured using a yaml file, how third-party tools can be executed efficiently across a cluster or on a local machine, and how data can be easily loaded into a database. Furthermore, we and others have been extensively using CGAT-core to build pipelines for computational genomics (https://github.com/cgat-developers/cgat-flow).\n\n\nDiscussion\n\nCGAT-core13 extends the popular Python workflow engine CGAT-Ruffus by adding desirable features from a variety other workflow systems to form an extremely simple, flexible and scalable package. CGAT-core provides seamless high-performance computing cluster interaction and adds Conda environment integration for the first time. In addition, our framework focuses on simplifying the pipeline development and testing process by providing convenience functions for parameterisation, database interaction, logging and pipeline interaction.\n\nThe ease of pipeline development enables CGAT-core to bridge the gap between exploratory data analysis and building production workflows. A guiding principle is that it should be as easy (or easier) to complete a series of tasks using a simple pipeline compared to using an interactive prompt, especially once cluster submission is considered. CGAT-core enables the production of analysis pipelines that can easily be run in multiple environments to facilitate sharing of code as part of the publication process. Thus, CGAT-core encourages a best-practice reproducible research approach by making it the path of least resistance. For example, exploratory analysis in Jupyter Notebooks can be converted to a Python script or used directly in the pipeline. Similarly, exploratory data analysis in R, or any other language, can easily be converted to a script that can be run by the pipeline. This lightweight wrapping of quickly prototyped analysis forms a lab book, enabling rapid reproduction of analyses and reuse of code for different data sets.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\n\nSoftware availability\n\nSource code available from: https://github.com/cgat-developers/cgat-core.\n\nArchived source code at time of publication: https://doi.org/10.5281/zenodo.259811513.\n\nLicence: MIT License.", "appendix": "Grant information\n\nThis work was funded by the Medical Research Council (UK) Computational Genomics Analysis and Training programme (G1000902).\n\n\nReferences\n\nAfgan E, Baker D, van den Beek M, et al.: The Galaxy platform for accessible, reproducible and collaborative biomedical analyses: 2016 update. Nucleic Acids Res. 2016; 44(W1): W3–W10. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWolstencroft K, Haines R, Fellows D, et al.: The Taverna workflow suite: designing and executing workflows of Web Services on the desktop, web or in the cloud. Nucleic Acids Res. 2013; 41(Web Server issue): W557–61. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOkonechnikov K, Golosova O, Fursov M, et al.: Unipro UGENE: a unified bioinformatics toolkit. Bioinformatics. 2012; 28(8): 1166–7. PubMed Abstract \| Publisher Full Text\n\nGolosova O, Henderson R, Vaskin Y, et al.: Unipro UGENE NGS pipelines and components for variant calling, RNA-seq and ChIP-seq data analyses. PeerJ. 2014; 2: e644. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNocq J, Celton M, Gendron P, et al.: Harnessing virtual machines to simplify next-generation DNA sequencing analysis. Bioinformatics. 2013; 29(17): 2075–83. PubMed Abstract \| Publisher Full Text\n\nGafni E, Luquette LJ, Lancaster AK, et al.: COSMOS: Python library for massively parallel workflows. Bioinformatics. 2014; 30(20): 2956–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVivian J, Rao AA, Nothaft FA, et al.: Toil enables reproducible, open source, big biomedical data analyses. Nat Biotechnol. 2017; 35(4): 314–316. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFisch KM, Meißner T, Gioia L, et al.: Omics Pipe: a community-based framework for reproducible multi-omics data analysis. Bioinformatics. 2015; 31(11): 1724–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKöster J, Rahmann S: Snakemake--a scalable bioinformatics workflow engine. Bioinformatics. 2012; 28(19): 2520–2. PubMed Abstract \| Publisher Full Text\n\nDi Tommaso P, Chatzou M, Floden EW, et al.: Nextflow enables reproducible computational workflows. Nat Biotechnol. 2017; 35(4): 316–319. PubMed Abstract \| Publisher Full Text\n\nLeipzig J: A review of bioinformatic pipeline frameworks. Brief Bioinform. 2017; 18(3): 530–536. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGoodstadt L: Ruffus: a lightweight Python library for computational pipelines. Bioinformatics. 2010; 26(21): 2778–9. PubMed Abstract \| Publisher Full Text\n\nHeger A, Cribbs A, Luna-Valero S, et al.: cgat-developers/cgat-core: First public release of code (Version v0.5.10). Zenodo. 2019. http://www.doi.org/10.5281/zenodo.2598115\n\nBray NL, Pimentel H, Melsted P, et al.: Near-optimal probabilistic RNA-seq quantification. Nat Biotechnol. 2016; 34(5): 525–7. PubMed Abstract \| Publisher Full Text\n\nLove MI, Huber W, Anders S: Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014; 15(12): 550. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "46758", "date": "16 Apr 2019", "name": "Devon P. Ryan", "expertise": [ "Reviewer Expertise Bioinformatics", "epigenetics", "immunobiology and neuroscience" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nWriting and using analysis pipelines has become a bioinformatician's (or more generally a data analyst's) bread and butter. There are a number of frameworks to perform such analyses, of which CGAT-Ruffus is preferred by many. CGAT-core brings some welcome functionality to Ruffus. In my mind, the addition of parameterisation and conda environment switching were the two biggest stumbling blocks to using Ruffus previously and CGAT-core seems to nicely address both of these.\nAfter going through the documentation (and a fair bit of trial and error due to not being particularly used to using Ruffus syntax) I was able to create and run a very simple workflow from scratch that included paired-end read trimming (cutadapt) and alignment (STAR) and used the parameterisation added by CGAT-core.\nWhile not critical to the manuscript, it'd be nice if the authors could address the following in the documentation (apologies to the authors if I missed some of these in the documentation):\nThe conda install instructions should be `conda install -c conda-forge -c bioconda cgatcore` Can examples of processing paired-end data be included in the showcase or elsewhere in the documentation? A particular step that I found difficult to implement the first time was performing read trimming such that the resulting files had the same name but were placed in a different directory. It turns out that this can be done with `formatter()`, but an example like the following in the documentation would probably be useful to new users like me:\n\n# pairs is a list of (read1, read2) tuples\n\n@transform(pairs, formatter(\".+/(?P.)_R[12].fastq.gz$\"), (\"trimmed/{SAMPLE[0]}_R1.fastq.gz\", \"trimmed/{SAMPLE[0]}_R2.fastq.gz\"))\nIs there any way to use standard commands to run jobs on a cluster rather than drmaa? Novice users are likely to have an easier time modifying `qsub` and `srun` commands than finding the drmaa shared libraries. At least in my testing it wasn't possible to use something like the following in a command:\n\ncmd = \"\"\"module load cutadapt\n\ncutadapt ...\"\"\"\nOne can combine the two commands with `&&` (or use a conda env), but I didn't notice this in the documentation. That's only a mild annoyance, but it should be mentioned to new users (especially those familiar with snakeMake, where commands are written to a shell script that's then submitted to the cluster).\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4723", "date": "16 Jul 2019", "name": "Adam Cribbs", "role": "Author Response", "response": "We would like to thank you for taking the time to review the manuscript, code and documentation so thoroughly. We are grateful for your helpful suggestions and we have updated our documentation in response as outlined below: 1. Writing and using analysis pipelines has become a bioinformatician's (or more generally a data analyst's) bread and butter. There are a number of frameworks to perform such analyses, of which CGAT-Ruffus is preferred by many. CGAT-core brings some welcome functionality to Ruffus. In my mind, the addition of parameterisation and conda environment switching were the two biggest stumbling blocks to using Ruffus previously and CGAT-core seems to nicely address both of these. After going through the documentation (and a fair bit of trial and error due to not being particularly used to using Ruffus syntax) I was able to create and run a very simple workflow from scratch that included paired-end read trimming (cutadapt) and alignment (STAR) and used the parameterisation added by CGAT-core. While not critical to the manuscript, it'd be nice if the authors could address the following in the documentation (apologies to the authors if I missed some of these in the documentation): The conda install instructions should be `conda install -c conda-forge -c bioconda cgatcore` Thank you, we have modified this in the documentation (https://cgat-core.readthedocs.io/en/latest/getting_started/Installation.html). 2. Can examples of processing paired-end data be included in the showcase or elsewhere in the documentation? A particular step that I found difficult to implement the first time was performing read trimming such that the resulting files had the same name but were placed in a different directory. It turns out that this can be done with `formatter()`, but an example like the following in the documentation would probably be useful to new users like me: # pairs is a list of (read1, read2) tuples @transform(pairs, formatter(\".+/(?P.)_R[12].fastq.gz$\"), (\"trimmed/{SAMPLE[0]}_R1.fastq.gz\", \"trimmed/{SAMPLE[0]}_R2.fastq.gz\")) Thank you, we have included this in the documentation on how to write pipelines (https://cgat-core.readthedocs.io/en/latest/defining_workflow/Writing_workflow.html#useful-information-regarding-decorators). 3. Is there any way to use standard commands to run jobs on a cluster rather than drmaa? Novice users are likely to have an easier time modifying `qsub` and `srun` commands than finding the drmaa shared libraries. It is possible to run standard job submission commands like qsub etc. as a command-line statement and run the pipelines in –no-cluster mode. We have added this to the documentation (https://cgat-core.readthedocs.io/en/latest/getting_started/Examples.html#using-qsub-commands). Furthermore, we have added some instructions in the installation page of the documentation of how to find drmaa libraries and set up an appropriate environment variable to store the location (https://cgat-core.readthedocs.io/en/latest/getting_started/Installation.html). 4. At least in my testing it wasn't possible to use something like the following in a command: cmd = \"\"\"module load cutadapt cutadapt ...\"\"\" One can combine the two commands with `&&` (or use a conda env), but I didn't notice this in the documentation. That's only a mild annoyance, but it should be mentioned to new users (especially those familiar with snakeMake, where commands are written to a shell script that's then submitted to the cluster). Apologies, the documentation did make this clear. The commands are written to shell scripts, but line breaks are not preserved. Hence: cmd = “””module cutadapt && cutadadapt … “”” will work as well as: cmd = “””module cutadapt; cutadapt ... “””” We have updated the documentation accordingly (https://cgat-core.readthedocs.io/en/latest/defining_workflow/Writing_workflow.html?highlight=module%20cutadapt#combining-commands-together )." } ] }, { "id": "47003", "date": "24 Apr 2019", "name": "Alexander Peltzer", "expertise": [ "Reviewer Expertise Evolutionary Biology", "Bioinformatics", "Workflow/Pipeline Development", "Systems Integration", "Human Genetics", "Population Genetics." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nCribbs et al describe CGAT-Core, a python framework for building scalable, reproducible computational biology workflows in their proposed software tool article.\nThe rationale behind the requirements for developing the software tool is explained properly, although there are many competitor tools and alternatives already \"on the market\" that can do similar or more things in general. The authors briefly summarize the large variance in portability, scalability, parameter handling and extensibility of the various tools in a sufficient form. One statement I cannot confirm in the last part of the introduction \"... in addition to python being an already widely understood language in computational biology, is that individual steps can use arbitrary python code, both in how they are linked together and in the actual processing tasks\". This is by all means not a drawback of a DSL, as some of these (e.g. nextflow, but to my knowledge also other competitors) allow for the direct integration of Python code in their respective tasks as well: https://www.nextflow.io/docs/latest/process.html#native-execution.\nThe statement in the methods section: \"Thus, parameters can be set specifically for each datasets, without the need to modify the code\" is a statement true for all workflow languages I know about (at least Snakemake, Nextflow and Toil separate configuration from actual pipeline code and can use e.g. parameter input files similar to the one that CGAT-Core uses, though with slightly different notation of course). As such, this statement should be probably changed or removed as it incorrectly makes readers assume that this is a unique feature of CGAT-Core. One could for example state that this is a best-practice pattern across various workflow languages and CGAT-Core also enables users to separate pipeline code from e.g. infrastructure or parameter descriptions.\nOne larger lack I see is that CGAT-Core unlike e.g. Snakemake, CWL, Nextflow and others does not support container technologies such as Docker or Singularity (to only name the two biggest solutions per market share out there). There have been papers critically investigating the effects of non-containerized conda environments \"which are ideal for packaging\" but not for long and mid-term reproducibility of analysis questions due to changing environments on a client side (see Grüning et al1 for details). These issues are only properly addressed by using containerization approaches, which CGAT-Core at the moment seems not to address. I think the authors should mention that the containerization of pipeline dependencies is a crucial part of reproducible data analysis today, and maybe whether they intend to add this in an upcoming release of CGAT-Core.\nTo not only mention critical parts: I do think the authors did a really good job in documentation, proper GitHub organization set up with README and all required information as well as setting up a community, which I'd like to congratulate them for.\nI was also able to run the mentioned example pipeline with Kallisto on my local infrastructure, although I had to adapt certain parts in order to be able to do so.\nI'm missing two points in the discussion: benefits for users to use CGAT-Core in general over other competitor tools, and also a more critical discussion on where the framework could be extended to support other environments (e.g. no mention of any cloud service/provider?) in general.\nOverall the paper reads well and I think it only requires minor changes, especially highlighting differences to other available workflow tools and critically assessing pros/cons with respect to these.\nTypos:\nDiscussion: \"... by adding desirable features from a variety other\" (missing \"of\")\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4724", "date": "16 Jul 2019", "name": "Adam Cribbs", "role": "Author Response", "response": "We would like to thank you for taking the time to review the manuscript. We are grateful for your perceptive suggestions and we have updated the manuscript, code and documentation in response as outlined below: 1. Cribbs et al describe CGAT-Core, a python framework for building scalable, reproducible computational biology workflows in their proposed software tool article. The rationale behind the requirements for developing the software tool is explained properly, although there are many competitor tools and alternatives already \"on the market\" that can do similar or more things in general. The authors briefly summarize the large variance in portability, scalability, parameter handling and extensibility of the various tools in a sufficient form. One statement I cannot confirm in the last part of the introduction \"... in addition to python being an already widely understood language in computational biology, is that individual steps can use arbitrary python code, both in how they are linked together and in the actual processing tasks\". This is by all means not a drawback of a DSL, as some of these (e.g. nextflow, but to my knowledge also other competitors) allow for the direct integration of Python code in their respective tasks as well: https://www.nextflow.io/docs/latest/process.html#native-execution. We did not mean to suggest that incorporation of Python code is a feature unique to CGAT-core, but that the ability to use Python to link tasks is a strength. Users do not have to learn a separate language for building a workflow, thus there is a very simple evolution from writing your first linear python script, then putting actions into functions and then combining these into a workflow through Ruffus decorators. We have reworded this statement to make our intention clearer. “A key advantage of this is that Python code can be used to link individual steps, as well as in processing tasks.\" 2. The statement in the methods section: \"Thus, parameters can be set specifically for each datasets, without the need to modify the code\" is a statement true for all workflow languages I know about (at least Snakemake, Nextflow and Toil separate configuration from actual pipeline code and can use e.g. parameter input files similar to the one that CGAT-Core uses, though with slightly different notation of course). As such, this statement should be probably changed or removed as it incorrectly makes readers assume that this is a unique feature of CGAT-Core. One could for example state that this is a best-practice pattern across various workflow languages and CGAT-Core also enables users to separate pipeline code from e.g. infrastructure or parameter descriptions. This statement was meant to compare with CGATcore with existing Ruffus functionality not other workflow tools. We have changed the text to make this clear. “Thus, parameters can be set specifically for each dataset, without the need to modify the code, a feature seen in many other workflow management systems.” 3. One larger lack I see is that CGAT-Core unlike e.g. Snakemake, CWL, Nextflow and others does not support container technologies such as Docker or Singularity (to only name the two biggest solutions per market share out there). There have been papers critically investigating the effects of non-containerized conda environments \"which are ideal for packaging\" but not for long and mid-term reproducibility of analysis questions due to changing environments on a client side (see Grüning et al1 for details). These issues are only properly addressed by using containerization approaches, which CGAT-Core at the moment seems not to address. I think the authors should mention that the containerization of pipeline dependencies is a crucial part of reproducible data analysis today, and maybe whether they intend to add this in an upcoming release of CGAT-Core. This is indeed a very important issue. This work is in our development plan and recently we have added prototype containerisation using kubernetes and/or singularity. We mention this ongoing development in the discussion section. “CGAT-core is under active development by the CGAT-Developers GitHub community. Support for cloud storage interaction, containerisation and LSF cluster interaction are currently being developed.” 4. To not only mention critical parts: I do think the authors did a really good job in documentation, proper GitHub organization set up with README and all required information as well as setting up a community, which I'd like to congratulate them for. I was also able to run the mentioned example pipeline with Kallisto on my local infrastructure, although I had to adapt certain parts in order to be able to do so. Many thanks, we believe it is important to follow best practices in software development and maintenance. 5. I'm missing two points in the discussion: benefits for users to use CGAT-Core in general over other competitor tools, and also a more critical discussion on where the framework could be extended to support other environments (e.g. no mention of any cloud service/provider?) in general. Overall the paper reads well and I think it only requires minor changes, especially highlighting differences to other available workflow tools and critically assessing pros/cons with respect to these. In this manuscript we cite a review (Leipzig, 2017) that compares many leading workflow tools including Ruffus, rather than perform a detailed comparison of CGAT-core with other tools ourselves. The focus of this paper was to elaborate the improvements that CGAT-core adds to Ruffus, building on the strengths highlighted by Leipzig and adding useful features available in other workflow tools. However, we do agree that the framework could be extended to support other environments. Indeed, we have added support for cloud data storage using the Boto3 SDK for AWS S3 storage interaction and moto for Google Cloud Storage. We are planning to add support for LSF to enable CGAT-core to be used in the Genomics England computing environment. We now mention these new developments in our discussion section. “Support for cloud storage interaction, containerisation and LSF cluster interaction are currently being developed.” 6. Typos: Discussion: \"... by adding desirable features from a variety other\" (missing \"of\") Many thanks, this typo has now been corrected." } ] } ]	1	https://f1000research.com/articles/8-377

Subsets and Splits

No community queries yet

The top public SQL queries from the community will appear here once available.